UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA COMMERCIAL PRODUCTION OF DESSERT WINES M. A. JOSLYN and M. A. AMERINE BULLETIN 651 SEPTEMBER, 1941 UNIVERSITY OF CALIFORNIA BERKELEY, CALIFORNIA CONTENTS Page Introduction 3 Economic status of the California wine industry 4 Types of dessert wines and their composition 12 Composition 12 Types 21 Principles of dessert-wine making. . 33 Varieties 33 Vinification 39 Fortification 46 Aging 58 Winery design, equipment, and operation 61 Design 62 Equipment 65 Sanitation and maintenance 71 Directions for making red dessert wines 73 Harvesting and fermentation .... 74 Conduct of the fermentation .... 75 Methods for color and flavor extraction 75 Pressing 77 Fortification 78 Aging 80 Directions for making sweet, white dessert wines 81 Harvesting and fermenting 82 Special procedures for muscatel . . 84 Fortification 85 Aging 87 Directions for making sherry and other rancio -flavored wines .... 90 Making Spanish sherry wine .... 91 Use of Spanish sherry yeast out- side of Spain 101 California process of sherry making 103 Other rancio-flavored wines 108 Page Grape concentrate and caramel sirup Ill Grape concentrate 112 Caramel sirup 114 Directions for making vermouth and related products 115 History in the United States .... 116 Nature of the herbs 117 Sweet (Italian) vermouth 117 Dry (French) vermouth 124 Other herb-flavored wines 125 Clarification and stabilization .... 125 Stabilization 126 Filtration 129 Fining 130 Centrifuging 133 Pasteurization 133 Preparation for marketing 134 Blending 134 Tasting 138 Bottling 139 Bacterial diseases and other dis- orders of dessert wines 143 Microbiological spoilage 143 Nonbacterial disorders 147 Analyses 150 Alcohol 150 Total acid 156 Volatile acid 156 Extract 157 Eeducing sugars 158 Tannin and coloring matter 160 Aldehyde 160 Esters 161 Color 161 Iron determination 162 Copper determination 163 Special tests 166 Interpretation of results 167 Acknowledgment 168 Selected references for further reading 169 Index 181 COMMERCIAL PRODUCTION OF DESSERT WINES 1 M. A. JOSLYN 2 and M. A. AMERINE 3 INTRODUCTION In Bulletin 639* the authors have considered the production of table wines — those which are ordinarily served with meals. The present publi- cation includes the wines and related beverages derived from wines which are not usually used as table beverages. These wines are called "dessert" wines because of their predominant usage. 5 Their alcoholic content is increased by the addition of grape spirits during production and formerly therefore they were called "fortified" wines. The brandy added in their production serves essentially as a preservative of the natural grape sugars present in the wine. They usually retain some sugar so that they are popularly known as "sweet" wines in contradis- tinction to table wines, which are usually, but not always, free of sugar and are called "dry" wines. We prefer to utilize the nomenclature of "table" and "dessert" wine, since this corresponds with the usage of the wines and is free of the ambiguity inherent in the "dry" and "sweet" system of naming. The use of the term "fortified" is objectionable be- cause of the implication that such wines resemble spirits. Vermouths, which are made from a wine base flavored with various herbs, are included in the discussion. As stated previously in the bulletin on table wines, the production of 1 Eeceived for publication May 24, 1941. 2 Assistant Professor of Fruit Technology and Assistant Chemist in the Experiment Station. 3 Assistant Professor of Enology and Assistant Enologist in the Experiment Station. 4 Amerine, M. A., and M. A. Joslyn. Commercial production of table wines. Cali- fornia Agr. Exp. Sta. Bui. 639:1-143. 1940. 5 The use of the dual appellation "dessert and appetizer" for the before- and after- dinner wines is unnecessary, since "dessert" wine corresponds closely enough with usage. For previous use of the term "dessert wines," see : Heide, C. von der, and F. Schmitthenner. Der Wein, 350 p. F. Viewveg und Sohn, Braunschweig, Germany. 1922. Miiller, K. Weinbau-Lexikon. 1,015 p. Paul Parey, Berlin, Germany. 1930. The Italian term vini di lusso or roughly "luxury wines" is not satisfactory. But the Likorweine, liqueur wines, used by Griinhut is not a bad appellation. He also uses "dessert" wine, but in a somewhat broader sense than it is used here. (See: Griinhut, L. Die Begutachtung der Dessertweine. Zeitschrift fur Untersuchung der Nahrungs- und Genussmittel 26:498-524. 1913.) In France vins de liqueur is used for the high-alcohol sweet wines and vins de liquoreux for the medium-alcohol natural sweet wines, such as Sauternes. Vino generoso is the Spanish term used for wines contain- ing added brandy. [3] 4 University of California — Experiment Station dessert and appetizer wines introduces problems that are quite different from those presented by the wines of lower alcoholic content. Their production is very closely regulated and restricted by the United States Bureau of Internal Revenue, since they are subject to higher taxes than table wines. This is particularly true of the steps involving changes in alcoholic content and volume — the production of grape fortifying brandy, and the fortification of the wine — both of which are carried out under the supervision of the United States gauger. The period of fermentation is much shorter than for dry wines so that losses due to poor practice are minimized, and after fortification they are less subject to spoiling during aging and storage. This has led to the current opinion that less skill is required in making sound dessert wines than is necessary for dry table wines. The production of dessert wines of quality, however, requires great skill and the use of the finest varieties and best methods. These wines can, however, be subjected to pasteurization, oxidation, chilling, and other operations, carried out for the purpose of more rapidly mellowing or stabilizing them, with a less deleterious effect on flavor than is true of table wines. This has led to a considerable misuse of such practices in order to market these wines at a very early age, particularly in increasing the rancio flavor by overoxidation or caramel- ization. For proper aging, dessert wines, in general, must be stored for a longer period of time and often in smaller containers than table wines. The commercial production of fortifying brandy and beverage brandy will be considered in a later publication. 6 The present bulletin is based on investigations of the divisions of Fruit Products and Viticulture as well as on observations of sound commercial practice. The discussion is primarily related to production under commercial winery conditions. Those chiefly interested in practi- cal directions for making dessert wines should turn directly to page 73. ECONOMIC STATUS OF THE CALIFORNIA WINE INDUSTRY 78 The commercial wine industry of California began in Los Angeles 9 in the middle of the nineteenth century and grew continuously until the passage of the Eighteenth Amendment restricted trade in wine. 6 Joslyn, M. A., and M. A. Amerine. Commercial production of brandies. California Agr. Exp. Sta. Bui. 652. (In press.) 7 Prepared in collaboration with S. W. Shear, Associate Agricultural Economist in the Experiment Station and Associate Agricultural Economist on the Giannini Foundation, who supplied the statistical data used. 8 General references on this subject in addition to those given in specific footnotes in the section are listed on p. 169. 9 Leggett, H. B. The early history of the wine industry in California. Thesis for the degree of Master of Arts, University of California. 1939. (Typewritten.) Copy on file in the University of California Library, Berkeley. Bul. 651] Commercial Production of Dessert Wines According to Bioletti, 10 the production in 1857 amounted to but 150,000 gallons. By the early eighties it had reached a total of about 10,000,000 gallons. Table 1 shows that by 1890-1892 it averaged 17,367,000 gallons TABLE 1 California Commercial* Wine Production and Grapes Crushed for Wine, Averages 1890-1939 and Annual 1933-1940 Grapes crushed for winef Commercial wine production, nett Years beginning July 1 Total Table wines § Dessert wines Quantity Per cent of total / 2 8 4 5 Averages: 1890-1892 tons 126,247 158,622 248,770 311,457 397,100 309,410 442,500 690,600 386,000 499,000 846,000 459,000 853,000 617,000 678,000 898,000 thousand gallons 17,367 19,630 28,497 35,204 43,595 36,120 36,342 63,908 35,679 37,005 65,690 46,679 85,351 50,342 71,478 103,000 thousand gallons 15,210 13,826 17,188 20,968 24,434 22,463 15,352 16,528 19,627 11,077 11,677 11,979 28,049 14,761 16,174 23,000 thousand gallons 2,157 5,804 11,309 14,236 19,161 13,657 20,990 47,380 16,052 25,928 54,013 34,700 57,302 35,581 55,304 80,000 per cent 12.4 1893-1898 29.6 1899-1903 39.7 1904-1908 40 4 1909-1913 44.0 1914-1918 37.8 1933-1934 57.8 1935-1939 74.1 Annual: 1933 45 1934 70.1 1935 82.2 1936 74.3 1937 67.1 1938 70.7 1939 77.4 19401 77.7 * In addition to commercial production of wine, quantities of homemade table wine produced from California grapes averaged 33,840,000 gallons in 1933 and 1934 and 32,889,000 gallons in 1935-1939. (The United States standard gallon, or wine gallon, is used in all tables.) t Excludes grapes crushed for the production of commercial beverage brandy. t Data are for net production, which is determined after allowances are made for normal shrinkage losses, distillation, diversion to by-product use, and other purposes. § Data on champagne and other sparkling wines included 1890-1918 but excluded 1933-1940. 1 Rough preliminary 1940 estimates subject to considerable revision. Sources of data: Compiled by S. W. Shear, Univ. California Giannini Foundation of Agricultural Economics. Averages 1890-1918 data from: Shear, S. W., and Gerald G. Pearce. Supply and price trends in California wine-grape industry, Part 2. Univ. California Giannini Foundation. Mimeo. Rept. 34. tables 9, 35, and 42. 1934. Col. 1: Based on data in table 42, cols. 4 and 5, and table 35, col. 6. Cols. 2-4: Data in table 9, cols. 2, 5, and 8. 1933-1940 data from: Shear, S. W. Deciduous fruit statistics as of January, 1941. Univ. California Giannini Foundation. Mimeo. Rept. 76. Grape tables 4, 15, and 16. Col. 1: Data from grape table 4, col. 5, minus data for 1933-1937 in grape table 16, col. 8, and minus data on beverage brandy production for 1938-1940 from the Wine Institute as of March 19, 1941. Cols. 2-4: Grape table 15, cols. 4-6. 10 Bioletti, F. T. The wine-making industry of California. International Institute of Agriculture, Agricultural Intelligence and Plant Diseases Monthly Bul. 6 (2) :1- 13. 1915. 6 University of California — Experiment Station a year, of which about one eighth consisted of dessert wines. u The industry grew so rapidly during the next two decades that by 1909- 1913 it had reached a pre-Prohibition peak of production of 43,595,000 gallons a year, constituting about 85 per cent of national production. Dessert wines had increased in production so much more rapidly than had table wines that they had risen to about 44 per cent of the total, or slightly over 19,000,000 gallons. Nearly half of this output of dessert wines consisted of the port type, according to the official classification of the United States Commissioner of Internal Revenue, nearly one third of sherry, and the balance mostly muscatel and Angelica, together with small amounts of Malaga, Madeira, and Tokay. During 1909-1913, consumption of commercial wine in the United States averaged about 50,000,000 gallons, or 0.52 gallon per capita (see table 2). Foreign wine constituted 7,392,000 gallons, or about 15 per cent of this total, and California wines about 80 per cent. Most of the wine not originating in California was dry table wine, so that only about 39 per cent of the national consumption consisted of dessert wines. Of the average per-capita consumption of commercial wine in the United States during that period, about 0.32 gallon consisted of table wines and 0.20 gallon of dessert wines. Almost no homemade wine was produced and consumed in the United States before Prohibition. Prohibition, however, stimulated the pro- duction of homemade wine to such an extent that per-capita consumption of wine actually reached its peak during Prohibition, for several years during that period amounting to more than the 0.77 gallon average con- sumption of all wines during the years since Repeal — 1933-1939. Most of the homemade wine consumed during Prohibition was presumably natural table wine, containing about 13 per cent alcohol, some of which was probably obtained from added sugar. The estimates by Shear given in table 2 show that even since Repeal a considerable amount of this un- taxed wine has been produced. Annual consumption of such wines made from California grapes probably has averaged close to 1 quart per capita since Repeal, with relatively little deviation from year to year. Con- sumption of tax-paid commercial wine has, however, increased substan- tially since Repeal, which has resulted in a decrease in the proportion of homemade wine from 47 per cent of total wine consumption in 1934 to 29 per cent in 1939. Roughly 60 per cent of the consumption of all dry table wines in 1939 still appears to have consisted of untaxed home- made wine made from California grapes. The wine industry has become much more important in the United 11 See also : Shear, S. W., and G. G. Pearce. Supply and price trends in the California wine-grape industry. Univ. California Giannini Foundation. Mimeo. Eept. 34. table 9. 1934. TABLE 2 United States Apparent Consumption of Dessert and Table Wines,* Years Beginning July 1 ; Averages 1909-1913 and 1935-1939, and Annual 1933-1940 Year beginning July 1 Total Commer- cial and homemade Commer- cial Dessert, over 14 per cent alcohol, commer- cial Table, not over 14 per cent alcohol Total Commer- cial Homemade Total consumption Averages: 1909-1913 1935-1939 Annual: 1933 1934 1935 1936 1937 1938 1939 1940f thousand gallons 49,445 100, 160 52,146 70,916 86,027 95,338 98,895 99,753 120,785 125,511 thousand gallons 49,445 67,271 17,526 37,856 50,012 65,503 64,230 70,533 86,075 90,741 thousand gallons 44,606 10,973 24,491 32,958 42,775 41,350 46,493 59,453 63,045 thousand gallons 30,247 55,554 41,173 46,425 53,069 52,563 57,545 53.260 61,332 62,466 thousand gallons 30,247 22,665 6,553 13,365 17,054 22,728 22,880 24,040 26,622 27,696 thousand gallons 32,889 34,620 33,060 36,015 29,835 34,665 29,220 34,710 34,770 Per-capita consumption Averages: 1909-1913 1935-1939 Annual: 1933 1934 1935 1936... . 1937 1938 1939 1940t ... gallt 0.52 .77 .41 .56 .67 .74 .76 .76 .92 0.95 gallons 0.52 .52 .14 .30 .39 .51 .49 .54 .65 gallons 0.20 .34 19 .26 .33 .32 .36 .45 0.48 gallons 0.32 .43 .32 .37 .41 .41 .44 .40 .47 0.47 gallons 0.32 .18 .05 .11 .13 .18 .17 .18 .20 0.21 gallons 0.00 .25 .27 .26 .28 .23 .27 .22 .27 0.26 Percentage of total consumption Averages: 1909-1913 1935-1939 Annual: 1933 1934 1935 1936 1937 1938 1939 1940t.... per cent 100.0 100 100.0 67.2 100.0 33.6 100.0 53.4 100.0 58.1 100.0 68.7 100.0 64.9 100.0 70.7 100.0 71.3 100.0 72.3 per cent per cent 38.8 44.6 21.0 34.5 38.3 44.9 41.8 46.6 49.2 50.2 per cent 61.2 55.4 79.0 65.5 61.7 55.1 58.2 53.4 50.8 49.8 per cent 61.2 22.6 12.6 18.9 19.8 23.8 23.1 24.1 22.1 22.1 per cent 32.8 66.4 46.6 41.9 31.3 35.1 29.3 28.7 27.7 * Includes imports which averaged 3,221,000 gallons a year during 1935-1939 or 3.2 per cent of consump- tion, about half dry and half sweet. t Preliminary estimates. Sources of data: Compiled by S. W. Shear, Univ. California Giannini Foundation of Agricultural Economics. Average 1909-1913 data from: Shear, S. W., and G. G. Pearce. Supply and price trends in the Cali- fornia wine-grape industry, Part 2. Univ. California Giannini Foundation. Mimeo. Rept. 34. table 7. 1934. 1933-1940 data from: Shear, S. W. Deciduous fruit statistics as of January, 1941. Univ. California Giannini Foundation Mimeo. Rept. 76:72. Grape table 14. 1941. 8 University of California — Experiment Station States since Repeal than in pre-Prohibition days. Table 1 shows that California production of commercial wines alone averaged nearly 64,000,000 gallons during 1935-1939, or nearly 50 per cent greater than during 1909-1913. Table 2 shows that United States consumption of all wine rose rapidly after Repeal from 0.56 gallon per capita in 1934 to 0.92 in 1939, averaging 0.77 gallon per capita during 1935- 1939. Only 3.2 per cent of the total was imported during 1935-1939. California consumed an average of about 23 per cent of the total national consumption during 1935-1939, or 3.58 gallons per capita. As in pre-Prohibition days, the consumption of dessert wines in- creased rapidly, rising from 0.19 gallon in 1934 to 0.45 gallon in 1939, and averaging 0.34 gallon per capita during 1935-1939. About 66 per cent of United States consumption of commercial wines consisted of dessert wines during 1935-1939, as compared with 39 per cent during 1909-1913. But when estimates of noncommercial, homemade, or base- ment wine are included in the total, only 45 per cent of all wine con- sumed in the United States appears to have been dessert wine during 1935-1939. There are no comprehensive statistics on the production of the different kinds of dessert wine since Repeal, such as were officially reported by the Commissioner of Internal Revenue before Prohibition. Vintners estimate, however, that port and sherry account for the greater proportion of the production, with muscatel, Angelica, Tokay, Madeira, and Marsala following in the order of importance. The latter two types are of very restricted importance. The geographical distribution of the California acreage of grapes for wine making is shown in table 3, while table 4 shows wine production for 1940 by chief counties and districts. These data show that the wine industry is widely distributed in the state. Although wines of all types are produced in nearly every district, production of dessert wines is very largely concentrated in the interior valleys, from Sacramento County south to Kern County, with the largest centers in Fresno and Lodi. Yields per acre are higher and cost of producing grapes lower in these great interior valleys than in the coastal valleys of the state, which very largely produce table wines. Some dessert wines are produced in the coast counties, however, largely from grapes shipped in from the interior valleys, and in part from grapes grown there. A limited but important production of dessert wines is centered in southern Cali- fornia. Muscatel wine is made largely in the lower San Joaquin Valley, where the acreage is largest, although large amounts of muscat grapes are delivered to wineries in other districts. There has been a tendency in recent years to divert more grapes of raisin and table varieties into wine and brandy than in pre-Prohibition Bul. 651] Commercial Production of Dessert Wines days. The fact that both raisin and table grapes have been in greater abundance and usually cheaper than before Prohibition has greatly increased their use by vintners. Moreover, winery demand for muscat TABLE 3 California Total Acreage* of Grapes by Classes and Wine Grapes by Varieties and by Districts,! 1936, and State Total, 1939 1939 1936 Variety and class State total State total t North coast Southern Calif- fornia Sacra- mento Valley Central Valley San Joaquin Valley 1 2 3 4 5 6 7 Totals: All varieties acres 514,414 254,401 85,019 174,994 159,445 53,307 30,854 29,321 10,971 8,143 7,819 3,269 15,761 15,549 3,996 2,981 1,603 506 6,463 acres 497,021 243,500 81,424 172,097 158,182 53,343 30,729 30,240 10, 164 7,977 7,508 2,980 15,241 13,915 3,020 2,639 1,545 507 6,204 acres 63,517 555 1,123 61,839 54,389 19,944 11,381 4,457 1,009 2,582 5,818 837 8,361 7,450 1,405 914 1,413 459 3,259 acres 43,914 10,431 4,555 28,928 26,512 7,706 2,387 3,470 4,352 3,506 93 1,222 3,776 2,416 599 1,070 61 15 671 acres 8,853 2,865 849 5,139 5,000 1,554 456 849 576 1,264 101 49 151 139 9 8 14 2 106 acres 84,068 7,268 28,300 48,500 47,156 18,335 12,123 11,919 2,645 382 892 150 710 1,344 567 404 26 1 346 acres 293,991 Raisin varieties 222,247 Table varieties 46,442 Wine varieties 25,302 Red wine, total 22,858 4,779 Carignane Alicante Bouschet Mission Mataro Petite Sirah . . . 4,281 9,434 910 216 599 Grenache Others White wine, total Palomino 722 1,917 2,444 440 Burger 207 Sauvignon vert 27 29 Others 1,741 * Includes total bearing and nonbearing acreage. t The districts (after Bioletti) include the following counties: North Coast — Lake, Marin, Solano, Mendocino, Napa, Sonoma, Alameda, Contra Costa, San Benito, Santa Clara, San Luis Obispo, and Santa Cruz; southern California— Riverside, San Bernardino, Imperial, Los Angeles, Orange, San Diego, and Ventura; Sacramento Valley — Placer, Sutter, Butte, Colusa, Glenn, Tehama, Yolo, and Yuba; Central Valley— Sacramento, San Joaquin, and Stanislaus; San Joaquin Valley — Fresno, Kern, Kings, Madera, Merced, and Tulare. t Includes sum of data in cols. 3-7 plus data in other counties not included in the district totals given. Sources of data: Compiled by S. W. Shear, Univ. California Giannini Foundation of Agricultural Economics, based on data from : Blair, R. E., W. R. Schreiber, and C. N. Guellow. California fruit and nut acreage survey 1936. U. S. Agricultural Adjustment Administration Statistical Publication 1: 11, 49, and 57-71. January, 1938. Blair R. E., and H. C. Phillips. Acreage estimates of California fruit and nut crops as of 1939. p. 24, 25. California Cooperative Crop Reporting Service, Sacramento, Calif. 1940. grapes has rapidly increased because of the greater popularity of mus- catel wine. Increased use of neutral wine as an alcoholic vehicle for pharmaceutical and similar beverage industries has also increased vint- ners' demand for the ordinary wines produced from raisin- and table- TABLE 4 California Commeecial Grape Crush, Gross Production of Wine, and Storage Capacity by Districts and Counties, 1940 District and county State total North coast, total North of San Francisco Bay: Mendocino Napa Sonoma Other§ South of San Francisco Bay: Alameda Contra Costa San Francisco Santa Clara Other§ Southern California:. . . Los Angeles San Bernardino . . . San Diego Other§ Sacramento Valley § Central Valley, total .... Sacramento San Joaquin Stanislaus San Joaquin Valley: Fresno Kern Kings Tulare Other§ Grapes crushed* for wine and brandy tons 995,981 138,986 106,532 5,725 32,483 65,520 2,804 32,454 8,516 733 44 22,467 694 76,841 12,037 63,147 968 689 2,743 288,195 42,655 204,276 41,264 489.216 281,323 30,827 12,002 132,276 32,788 Wine production, gross t Total thousand gallons 105,690 22,004 17,431 963 5,395 10,727 4,573 1,182 129 7 3,146 109 8,740 1,496 7,014 129 101 362 29,780 4,239 21,301 4,240 44,804 24,607 2,753 1,752 12,928 2,764 thousand gallons 78,320 3,424 1,860 122 328 1,212 1,564 413 1,150 1 6,603 1,025 5,517 46 15 145 25,112 4,055 18,546 2,511 43,036 23,893 2,735 1,697 12,283 2,428 Table Red thousand gallons 20,362 15,047 13,039 733 3,999 8,229 78 2,008 330 87 7 1,511 73 1,433 270 1,023 65 75 192 2,938 125 1,335 1,478 752 255 14 247 236 White thousand gallons 7,008 3,533 2,532 108 1,068 1,286 70 1,001 439 42 485 35 704 201 474 18 11 25 1,730 59 1,420 251 1,016 459 4 55 398 100 Storage capacity! thousand gallons 196,235 56,556 41,970 2,346 13,475 25,106 1,043 14,586 3,841 524 3.453 6,393 375 20,035 4,230 14,950 503 352 613 34,405 7,730 70,067 42,922 3,389 1,018 16,735 6,003 * Data are for July 1-December 31, 1940. Small quantities of raisins and other fruits not included. t Gross wine production as of December 31, 1940, without allowances for subsequent losses, removals for distillation, increases resulting from amelioration and fortification, etc. t Storage capacity as of December 31, 1940. Includes fermenters usable for storage. § District data designated as "other" may include small amounts outside of the district. "Other" includes the following counties by districts: North of San Francisco Bay — Humboldt, Lake, Marin and Solano; South of San Francisco Bay — Monterey, San Benito, San Mateo, and Santa Cruz; southern California— Riverside, Santa Barbara, and Ventura; San Joaquin Valley — Madera and San Luis Obispo; Sacramento Valley, total, includes Amador, Butte, Calaveras, Merced, Nevada, Yolo, Yuba, and Placer. Source of data: Compiled by S. W. Shear. Uni v. California Giannini Foundation of Agricultural Economics, from: Wine Institute. Fifth wine industry statistical survey. Confidential Bui. 102: 18, 19. March 19, 1941. (Mimeo.) Bul. 651] Commercial Production op Dessert Wines 11 grape varieties, such as Thompson Seedless (Sultanina) and Emperor. These grapes are considered useful, however, in the production of neutral grape spirits for fortifying dessert and appetizer wines and for sale as a fruit spirit. Since Repeal almost no California grapes have been wasted, the wine and brandy industry utilizing all that have not been shipped fresh or dried. Of the average annual quantity of California grapes harvested during 1935-1939, about 44 per cent was used for commercial and non- commercial wine and brandy, or approximately 986,000 tons. Of this total, about 576,000 tons, or 58 per cent, consisted of wine-grape varie- ties, about 227,000 tons, or 23 per cent, of raisin varieties, and 183,000 tons, or 19 per cent, of table varieties. More California grapes were used for wine and brandy making in 1940 than ever before, preliminary esti- mates indicating that a total of about 1,186,000 tons, or 54 per cent of the 1940 crop, was used by vintners. About 51 per cent of this total con- sisted of wine varieties, 33 per cent of raisin, and 16 per cent of table varieties. The increasing supply of wine made from raisin and table grapes and the lower quality of much of this wine have greatly increased the marketing problems of the wine industry. The need of increasing the demand and consumption for such wine is accentuated by the un- usually large quantities of raisin and table grapes diverted into wine and brandy making from the 1940 crop as a result of the loss of much of the market for California raisins and table grapes, particularly the export market for raisins. Preliminary estimates indicate that about 103,000,000 gallons of commercial wine were produced in California in 1940, or about 31,500,000 gallons more than in 1939. Imports of foreign wines into the United States averaged 3,221,000 gallons a year during 1935-1939. The marked decrease in imports resulting from the European war in 1940 and 1941 caused some increase in the demand for California wines. Export markets for wine were investigated in this period, and some increase in the small amounts formerly exported will probably occur if the war continues to restrict exports from the European countries. Vermouth, made by blending basic fortified wines with characteristic flavoring materials, is being made in the United States in increasing quantity, as shown in table 5. Liberalization of restrictions governing its manufacture, decrease in taxation, and reduction in imports as a result of the European war of 1939 have accounted for the rapidly growing vermouth industry. California wines are used, almost ex- clusively, as bases for United States vermouth; the flavoring materials, however, are largely imported. 12 University of California — Experiment Station TYPES OF DESSERT WINES AND THEIR COMPOSITION 12 COMPOSITION The chemical composition of dessert wines differs in several important respects from that of table wines. These differences are due to several factors, of which the most important are those resulting from differ- TABLE 5 United States Production and Consumption of Vermouth, Years Beginning July 1. 1934-1939 Production United States consumption Year begin- United States Cali- fornia *t Total Foreign Domestic ning July 1 Total At wineriesf At rectifying plants! Quantity Per cent of total / 2 3 4 5 6 7 8 1934 1935 1936 1937 1938 1939 gallons 138,658 180,527 243,444 220, 192 216,984 488,785 gallons 164,747 201,481 206,184 479,074 gallons 138,658 180,527 78,697 18,711 10,800 9,711 gallons 25,911 40,658 71,568 105,536 78,249 156,181 gallons 1,070,642 1,125,748 1,463,013 1,325,412 1,436,716 2,185,461 gallons 931,984 945,221 1,301,604 1,153,494 1,238,628 1,781,505 gallons 138,658 180,527 161,409 171,918 198,088 403,956 per cent 13.0 16.0 11.0 13.0 13.8 18.5 * California production at rectifying plants excluded for 1937-1939 because data not available. Vermouth produced at California rectifying plants in wine gallons was: 1934, 25,911; 1935, 40,658; and 1936, 36,647; and was probably very small since 1936 judging by United States production in column 3. t Production of vermouth in wineries was first permitted June 26, 1936, under the Liquor Tax Adminis- tration Act. t Vermouth produced at rectifying plants officially reported in proof gallons, and here multiplied by 2.77777 to convert to approximate wine gallons. Source of data: Compiled by S. W. Shear, Univ. California Giannini Foundation of Agricultural Economics. Col. 1: Sum of cols. 2 and 3. Col. 2: From: U. S. Bureau of Internal Revenue, Alcohol Tax Unit. Statistics on wine. Annual mimeographed reports. Col. 3: 1934-1936 from: United States Tariff Commission. Grapes, raisins and wines. U. S. Tariff Comm. Rept. 2d ser. 134: 242. 1937-1939 from: U. S. Bureau of Internal Revenue, Alcohol Tax Unit. Statistics on distilled spirits and rectified spirits and wines. Annual mimeographed reports. Col. 4: Same source as cols. 2 and 3. Col. 5: Sum of cols. 6 and 7. Col. 6: July-Dec, 1934, estimated. Jan., 1935-Dec, 1936, from: Wine Institute. Third wine industry statistical survey. Bui. 86: 33, 34. July 19, 1939. Jan., 1937-June, 1940, from: U. S. Bureau Foreign and Domestic Commerce, Imports of distilled liquors, wines, cordials, and malt liquors. Mimeographed Monthly Release 3063. Col. 7: Sum of United States tax-paid withdrawals of vermouth from wineries, as reported by: U. S. Bureau of Internal Revenue, Alcohol Tax Unit. Statistics on wine. Annual mimeographed reports. Plus production in col. 3. Tax-paid withdrawals of vermouth made at rectifying plants are not available in official reports. ences in production and handling; namely, (a) the addition of fortifying brandy to dessert wines, with the resulting dilution of the fermentation products as well as other constituents and secondary effects of the higher alcohol content, for example, precipitation of tartrates and extraction of 12 General references on this subject in addition to those given in specific footnotes in this section are listed on p. 169-71. Bul. 651] Commercial Production of Dessert Wines 13 substances from the wood; (6) the restricted period of fermentation of dessert wines and the resulting decrease in fermentation products and in extraction of certain substances from the skins; (c) the heating of dessert wines or the addition of heated musts or wines to them; (d) the different chemical composition of the raw material; (e) the increased period of aging accorded some dessert wines — usually in moderately warm cellars. The influence of each of these factors on the composition of dessert wines will be more apparent in the discussion of the individual constitutents which follows. Acids. — The varieties of grapes used for dessert wines are naturally lower in total acid content and higher in pH than those used for the table wines, and the sugar-acid ratio is greatly increased. 13 Since the grapes intended for dessert wines are commonly grown in the warmer districts of California, their acidity is still further restricted. The dilu- tion effect of fortification is also an important factor in the reduced acidity of the finished wine. Finally, the higher alcohol content after fortification favors the precipitation of tartrates. The amounts of suc- cinic acid present in dessert wines is also less than that of table wines owing to the reduced period of fermentation. Acid tartrate, but usually not free tartaric acid, is the chief acid material present. Small amounts of malic acid and acid malates are also found. Citric acid is reported in sweet wines in amounts as low as those commonly found in table wines, or lower. 14 California dessert wines are generally lower in total acid than com- parable European types. The reduced acidity is so exaggerated in some musts as to make them difficult to ferment cleanly as well as to make the resulting wines taste too flat. These low acidities are partly due to the very large amounts of raisin and table grapes used for dessert wines in California — varieties which, in general, have very low acidities; partly to the delayed harvesting practices current in California ; and partly to the excessive crops. A range of 0.30 to 0.65 per cent total acidity for finished dessert wines is sufficiently broad and a smaller range of 0.40 to 0.65 may be desirable. At present, acidities as low as 0.20 per cent are found in some California dessert wines. (See tables 8-12.) Acetic acid is found in only small amounts in dessert wines unless spoilage starts in the must in the early stages of fermentation, as some- 13 Amerine, M. A., and A. J. Winkler. Maturity studies with California grapes. I. The Balling-acid ratio of wine grapes. American Society for Horticultural Science Proceedings 38:379-87. 1941. 14 Heiduschka, A., and C. Pyriki. Beitrag zur Kenntnis des Citronensauregehaltes von Traubenmosten und Traubenweinen. Zeitschrift fur Untersuchung der Leben- smittel 54:466-73. 1927. 14 University of California — Experiment Station times happens. 15 Only an occasional post-Prohibition California dessert wine, however, shows a volatile acid content which is above the Cali- fornia limits of 0.110 per cent (0.120 per cent outside of California). (See tables 6 and 8-12.) The pH values of California dessert wines are much higher than those of California table wines. They range from about 3.5 to 4.2, as compared with 2.9 to about 3.8 for table wines. These high pH values result from the reduced acidity and restricted buffer capacity of dessert-wine musts as well as from the influence of their increased alcohol and sugar con- tents. TABLE 6 Limits for Certain Constituents in Dessert Wines Set by the California State Department of Public Health and by the Federal Regulations Authority Alcohol,* range Maximum volatile acid as acetic Minimum sugar per cent 19.5-21.0 18. 0-21. Of grams per 100 cc 0.110 0.120§ Balling degree 5 5f * Sacramental wines are exempted. The general practice of the government authorities is to permit a 0.5 per cent tolerance. t Except for sherry which now has a maximum of 4 per cent reducing sugar. At a recent meeting (July, 1941), while this bulletin was in press, the Standardization Committee of the Wine Institute met and proposed that the reducing sugar content of wines labeled "California dry sherry" be to V/ 2 per cent, those labeled "Calif or nia sherry, " to 4 per cent, and those labeled "California sweet sherry," 4 to 7 per cent. The minimum Balling degree for California Tokay was set at 3.5°. A minimum total acidity for all dessert wines of 0.30 per cent was proposed. t Vermouth has a minimum limit of 15 per cent; sherry has a minimum of 17 per cent. § Exclusive of sulfur dioxide. Alcohols. — The alcohol content of dessert wines may vary from 18.5 to 21.0 per cent according to federal law, and in California the limits are from 19.5 to 21.0. These limits are closer than those found in com- parable European types. The enforcement of the 19.5 minimum limit also forces the fortification of some dessert wines (uncooked sherries, for example) to an undesirably high level (21.0 per cent). At a lower original percentage of alcohol, the wines would become palatable sooner and the dilution of the aroma and extract by the fortification would be reduced. Reduction of the minimum limit to about 17.5 per cent would therefore be of considerable value in improving quality, but might en- courage dilution of dessert with table wines. Ethyl alcohol is the chief alcohol present, but measurable quantities of other alcohols are also found, mainly methyl, amyl, isoamyl, w-butyl, iso- butyl, w-propyl, hexyl, and heptyl. Since most of the alcohols in dessert wines — with the possible exception of some Spanish sherries — are de- rived from the fortifying brandy, they will be discussed in Bulletin 652 15 Cruess, W. V. "Volatile" in Muscatel. The Wine Eeview 4 (5) : 18-20. 1937. Bul. 651] Commercial Production of Dessert Wines 15 (cited in footnote 6, p. 4). The total amount of the higher alcohols pres- ent does not exceed 1 per cent of the total alcohol content. The methyl alcohol content is also low. Methyl alcohol may, however, be expected in dessert wines made with fortifying brandy high in this substance. Mannite, a polyhydric alcohol, is formed in wines as a result of bac- terial action. In European wines it commonly occurs as a result of infection during fermentation, but under California conditions it is formed chiefly as a result of infection during storage (see p. 144). It is not usually harmful to the taste of dessert wines but does indicate bac- terial contamination. During aging, losses in alcohol may occur by evaporation, directly or during routine winery operations, or by esterification. In Spain, under certain conditions, increases in alcohol during aging, amounting to several per cent, are reported. 16 This is apparently due to differential loss of water through the wood during* aging. Such changes are rare in California, where very large containers are used more commonly. Usually small losses are reported, particularly during the baking of sherry, during pasteurization, filtration, centrifuging, and racking. Sugars. — The two chief sugars present in grapes are dextrose and levulose. During ripening of the grapes, the levulose-dextrose ratio of the grapes increases to about 1. After full maturity the ratio may slightly exceed 1. Since dessert wines are made from fully ripe grapes, ratios of at least 1 are common. According to Hopkins and Roberts, 17 the dextrose is fermented more rapidly than levulose so that the levu- lose-dextrose ratio should increase during the early stages of fermenta- tion. Wines whose fermentation has been stopped by the addition of alcohol should then show a ratio considerably above 1, and Kniphorst and Kruisheer 18 have found this to be the case in genuine European port, Tokay, and Sauternes wines. 19 Both fortified musts and wines which have been fermented nearly dry and then sweetened with concentrate showed 16 Castella, F. de. Sherry. Victoria Department of Agriculture Journal 24:690-98. 1926. Gonzalez Gordon, M. M a . Jerez-Xeres-"Scheris." 405 p. Imprenta A. Padura, Jerez de la Frontera, Spain. 1935. Anonymous. The Lancet Analytical Commission on sherry: its production, com- position, and character. Lancet 1898 (II) : 1134-40. 1898. 17 Hopkins, E. H., and R. H. Eoberts. Kinetics of alcoholic fermentation of sugars by brewer's yeast. II. The relative rates of fermentation of glucose and fructose. Biochemical Journal 29:931-36. 1935. 18 Kniphorst, L. C. E., and C. I. Kruisheer. Die Bestimmung von 2-3 Butylenglykol, Acetylmethylcarbinol und Diacetyl in Wein und anderen Garungsprodukten. II. An- wendung des Verfahrens auf die Untersuchungen einiger Weintypen. Zeitschrift fur Untersuchungen der Lebensmittel 74:477-85. 1937. 19 Harden reports that the Sauternes strain of wine yeast ferments levulose more rapidly than dextrose. (Harden, Arthur. Alcoholic fermentation. 4th ed. 243 p.; see especially p. 192-94. Longmans, Green and Co., London. 1932.) 16 University of California — Experiment Station the expected ratio of about 1. These included certain wines of Malaga and Tarragon, Spain, which are frequently prepared by the use of reduced musts. Similar results are reported by Szabo and Rakcsanyi, 20 who found that when wine was made from grapes of a very high initial sugar concentration, the levulose-dextrose ratio was near 1 after fer- mentation. They also found the ratio to be about 1 when concentrate was added. But they found two to six times as much levulose as dextrose in natural sweet wines of low sugar content. Similar results have been obtained in the fermentation of a concentrate with Zygosaccharomyces sp. by Parisi, Sacchetti, and Bruini. 21 These workers also report that in very sweet musts the levulose fermented more rapidly than dextrose in the early stages of fermentation. Other changes in composition due to differences in the rate of fer- mentation of levulose and dextrose cause the originally dextrorotatory must to become levorotatory during the early stages of fermentation. Since dextrose is less sweet to the taste than levulose (in relation to sucrose as 100, the sweetness of dextrose is about 74 and that of levulose is about 173 22 ), there will be a difference in the sweetness of a sweet wine made from a partially fermented must as compared to a wine made by the addition of sucrose or grape concentrate to a dry wine, 23 the former being appreciably sweeter than the latter. The determination of the levulose-dextrose ratio is a rough means of determining the nature and extent of the fermentation period of dessert wines, if produced from musts of moderate sugar concentration. Muttelet 2 * was unable to find sucrose in genuine Portuguese port wines, and Clavera and Lopez 25 found none in similar samples of Malaga. There is abundant evidence that grapes of Vitis vinifera contain very little or no sucrose. Alwood 26 has shown that sucrose occurs in certain 20 Szabo, I., and L. Kakcsanyi. Das Mengenverhaltnis der Dextrose und der Lavu- lose in Weintrauben, im Most und im Wein. Ve Congres International Technique et Chimique des Industries Agricoles Comptes Eendus 1:936-49. 1937. 21 Parisi, E., M. Sacchetti, and C. Bruini. Sulla fermentazione alcoolica dei mosti concentrati. Annali di Chimica Applicata 22:616-20. 1932. 22 Biester, Alice, M. M. Wood, and C. S. Wahlin. Carbohydrate studies. I. The rela- tive sweetness of pure sugars. American Journal of Physiology 73 : 387-96. 1925. For other data see: Walton, C. F., Jr. Sweetening agents. Relative sweetening power, p. 357-58. In: Washburn, E. W. International critical tables. Vol. I. 415 p. McGraw-Hill Book Co., New York, N. Y. 1926. 23 But invert sugar, the product of the hydrolysis of sucrose and containing equal amounts of dextrose and levulose, has a relative sweetness of about 127-130. 24 Muttelet, C. F. Les sucres des vins de Ports. Annales des Falsifications et des Fraudes 23:205-7. 1930. 25 Clavera, J. M a ., and M. O. Lopez. Los azucares y el extracto seco en los vinos do Malaga. Sociedad Espafiola de Fisica y Quimica Anales 30': 140-44. 1932. 26 Alwood, W. B. Enological studies. U. S. Dept. Agr. Bur. Chem. Bui. 140:1-24. 1911. Bul. 651] Commercial Production of Dessert Wines 17 native American species, but even in wines made wholly from such grapes the percentage of sucrose will be small. Since genuine California wines are practically always made from pure V. vinifera varieties and since, furthermore, the addition of any sweetening agent except grape concentrate is prohibited by California law, the presence of hydrolyzable sugars, such as sucrose, in California sweet dessert wines, is very strong evidence indeed of adulteration. Kickton and Murdfield, 27 however, report a supposedly genuine Ma- deira wine containing over 3 per cent of sucrose. Practically all of their 240 samples, however, were free of sucrose. Kickton and Korn 28 also re- port sucrose in 13 out of 591 samples of sherries, all of which were pre- sumably genuine. Products Derived from Sugars. — When acid solutions containing hex- ose sugars are heated for a sufficiently long period or at a high enough temperature, a dehydration takes place and hydroxymethylfurfural , — o , (CH 2 0-C=CHCH=C-COH) is formed. Boiled-down musts, and heated wines, such as California sherries, contain considerable amounts of this substance. Methods of testing for hydroxymethylfurfural are given on page 166. Jagerschmid 29 has used one of these methods for the deter- mination of added caramel to brandies, while Kruisheer, Vorstman, and Kniphorst 30 and Botelho 31 have used the test to determine the adultera- tion of wines such as genuine Portuguese port, which are supposed to be made only with fresh grapes, with reduced musts or with caramel. Al- though hydroxymethlyfurfural is formed in honeys during storage, Kruisheer and co-workers did not find excessive quantities in a pure thirty-five-year-old wine. Further investigation of this point should be made, particularly of very sweet wines stored in small cooperage in warm cellars. Hydroxymethylfurfural has an agreeable odor like that of camomile. Its taste, however, is slightly bitter. 32 Other products which Kruisheer and co-workers found in heated 27 Kickton, A., and E. Murdfield. Herstellung, Zussammensetzung und Beurteilung des Madeiraweines und seiner Ersatzweine. Zeitschrift fur Untersuchung der Nah- rungs und Genussmittel 28:325-64. 1914. 28 Kickton, A., and 0. Korn. Herstellung, Zusammensetzung, und Beurteilung des Sherrys und seiner Ersatzweine. Zeitschrift fur Untersuchung der Nahrungs- und Genussmittel 47:281-328. 1924. 29 Jagerschmid, A. Nachweis von Karmel in Wein, Kognak und Bier. Zeitschrift fur Untersuchungen der Nahrungs- und Genussmittel 17:269. 1909. 30 Kruisheer, C. I., Vorstman, and L. C. E. Kniphorst. Bestimmung von Oxy- methylfurfurols und des Lavulosins in Portwein und anderen Sussweinen. Zeit- schrift fiir Untersuchungen der Lebensmittel 69:570-82. 1935. 31 Botelho, L. C. Dosage de l'oxymethylfurfurol dans le vin de Porto. Eevue de Viticulture 89:202-5. 1938. 32 Middendorp, J. A. Sur l'oxymethylfurfurol. Eecueil des Travaux Chimiques des Pays-Bas 38:1-71. 1919. 18 University of California — Experiment Station sweet wines or musts were levulinic acid (CH 3 C : OCH 2 CH 2 COOH) and substances called Lavulosins. These are apparently dehydration products from levulose. 33 Other investigators have been able to detect the use of raisins in wines by determining the fluorescence of the wines under ultraviolet light. 3 * 2, 3-Butylene Glycol. — Although 2, 3-butlyene glycol does not seem to have any direct influence on the taste and aroma of wines (in pure solution it is practically odorless), its presence is of interest because the amount present seems to be correlated with the extent of the fermenta- tion. 35 For this reason, wines fortified before the completion of their fermentation should contain reduced amounts of this substance, and such is the case. Kniphorst and Kruisheer found 40 to 80 mg of it for each per cent of alcohol in dry wines and only 3 to 15 mg for each per cent of alcohol in fortified wines. They did not however, find any acetyl- methylcarbinol or diacetyl — the two successive stages in the oxidation of 2, 3-butylene glycol — in wines. Garino-Canina 36 also found reduced amounts of 2, 3-butylene glycol and confirms the fact that it is produced in proportion to the extent of fermentation in wines of low alcohol and in fortified wines. He has found acetylmethylcarbinol in wines fer- mented in the presence of acetaldehyde, and large amounts in vinegars. Parisi, Sacchetti, and Bruini (cited in footnote 21, p. 16) also found acetylmethylcarbinol in the products of a fermentation of grape con- centrate with a Zygosaccharomyces. Glycerin. — Since glycerin is a by-product of fermentation, the amounts of it found in dessert wines would be expected to be small owing to the reduced period of fermentation and to the dilution factor of fortification. Such is the case. The amounts found are smaller than those of most table wines, although unexpectedly large amounts are found in some dessert wines. The range reported is from 0.03 to 1.5 per cent. The sweetness of dessert wines and their generally high specific grav- ity makes the influence of glycerin on the taste and texture of the wine less important than in table wines. In addition, since the restricted fer- 33 Gisvold, Ole, and C. H. Eogers. The chemistry of plant constituents. 309 p. (See especially p. 41-42.) Burgess Publishing Co., Minneapolis, Minnesota. 1939. 34 Szabo, I. Examination of wine by means of a quartz lamp. [Translated title.] Magyar Ampelol. Evkonyr 9:454-57. 1935. Abstracted in: Chemical Abstracts 30: 1506. Canals, E., and H. Collet. Spectres de fluorescence des vins. Annales des Falsifica- tions et des Fraudes 32:163-71. 1939. 35 Joslyn, M. A. The by-products of alcoholic fermentation. Wallerstein Labora- tories Communications 3(8):30-43. 1940. 30 Garino-Canina, E. II 2-3 butilenglicole e l'acetilmetilcarbinolo nci vini e negli aeeti. Annali di Chimica Applicata 23:14-20. 1933. Bul. 651] Commercial Production of Dessert Wines 19 mentation and use of fortifying brandy during production is generally recognized for all California dessert wines, the determination of gly- cerin in these wines is of minor importance. Furthermore, the recogni- tion of the very marked influence of environmental conditions on the amount of glycerin formed has diminished its value in the detection of adulteration, even for table wines. Its concentration is, however, of value as an indication of the extent of fermentation. Aldehydes. — Although small amounts of acetaldehyde are normally produced during fermentation, the largest amounts are apparently formed during aging. Amounts from 10 to 200 mg per liter are found in various types of California dessert wines. Normally the largest amounts — 100 mg or over per liter — are found in sherries which have TABLE 7 Acetaldehyde in Various Types of California Dessert Wines Samples Volatile acidity Sulfur dioxide Acetaldehyde Type Range Aver- age Range Aver- age Range Aver- age California sherry California dry sherry . . number 29 19 per cent 0.026-0.122 0.026-0 096 per cent 0.059 0.058 p. p. TO. 9.6- 99.4 6.4-160.0 p. p. TO. 33.9 42.0 p. p. TO. 14.9- 99.9 17.7-179 p. p. TO. 55.4 78.8 undergone a film-yeast fermentation. But considerable amounts are also found in California sherries (table 7) prepared by the baking process. Acetals. — Trillat 37 found acetals to be present in wines and brandies. Acetals arise from the reaction of acetaldehyde and alcohols. They seem to be particularly important in wines of the sherry type, where the high content of aldehyde and alcohol favors their formation. The amounts found were small — 150-190 mg per liter of wine. Esters. — Numerous esters are found in wines. They are mainly ethyl esters. They range from the volatile esters of the low-molecular-weight acids, such as acetic, to the ethyl acid esters of acids, such as tartaric and malic. These latter esters have a boiling point above that of water. Esters are formed in wines through biological activity of yeasts and bacteria and through esterification during aging. According to the early enologists, all esters were believed to have a marked influence on the desirable odors of wines. 38 More recent work indicates that for table wines, at least, only ethyl acetate is an im- 37 Trillat, A. L'aldehyde acetique dans le vin; son origine et ses effets. Institut Pasteur [Paris] Annales 22:704-19, 753-62, 876-95. 1908. 38 Rocques, X. Le bouquet des vins. Revue de Viticulture 12:95-99. 1899. 20 University of California — Experiment Station portant factor in the aroma, and this mainly as a spoilage product. 30 In dessert wines the increased amount of alcohol should favor the in- creased production of all esters, and in fact aged dessert wines are markedly higher in esters. Rocques, for example, reported a very high concentration of esters in Madeira wines. As in table wines, the volatile neutral esters are undesirable, and if they amount to 200 mg per liter, the wine takes on a definitely spoiled character. Tannin. — The tannins in wine are derived from the grape during fermentation, particularly from the skins, from the wood during aging, and by direct addition. The tannin derived from the wood is of little importance except in the case of white wines : Ribereau-Gayon and Peynaud 40 have shown that white wines stored in barrels may gain in tannin. Garino-Canina 41 questions the importance of tannin from the grapes which are found in wines. He finds most of the reactions attrib- uted to tannins to be due to the pigments in the wine, especially to oenin chloride. More recently, Negre 42 has used the differential precipitation of total tanninlike material by lead acetate and the true tannin pre- cipitated by zinc acetate to determine the amount of nontannin poly- phenols present. He finds the latter are dissolved primarily in the early stages of fermentation while the oenotannins are dissolved by the alco- hols primarily in the latter stages of fermentation. The significance of these results in fortification practices has not yet been determined. The tannins have an important effect on the taste of the wine and also influence its stability. During the aging of wines, the tannin content gradually decreases, notably in red wines. The tannins are also important in wine because of their ability to precipitate proteins, and the tannin content is markedly reduced by fining agents such as gelatin. It is also slightly reduced by filtration through cellulose or asbestos. See page 59 for the chemical changes in red wines during aging. Coloring Material. — The stability of color in red wines is very im- portant. The color in Vitis vinifera grapes is due to anthocyanin pig- ments but no complete study of the nature of these pigments nor of the 39 Ribereau-Gayon, J., and E. Peynaud. Esterification chimique et biologique des acides organiques du vin. Societe Chimique de France Bui. 3(serie 5):2325-30. 1936. Peynaud, E. L'acetate d'ethyl dans les vins atteints d'acescence. Annales des Fermentation 2:367-84. 1936. 40 Ribereau-Gayon, J., and E. Peynaud. Etudes sur le collage des vins (II). Revue de Viticulture 81:53-59. 1934. 41 Garino-Canina, E. Contribution to the study and to the determination of the tannins and the pigments of the grape. [Translated title.] Stazioni Sperimentali Agrarie Italiane 57:245-74. 1924. 42 Negre, E. Contribution oenologique a l'etudes des matieres tannoides. L'Aca- demie d'Agriculture de France Comptes Rendus 25:647-52. 1939. Bul. 651] Commercial Production of Dessert Wines 21 influence of oxygen on them has been made. (See Bulletin 639, cited in footnote 4, p. 3.) TYPES No completely satisfactory system of nomenclature is available for California dessert wines. The use of foreign type names dates back to the time of the gold rush, although their continued use has been the subject of considerable criticism and polemics. The evolution of a system of naming which is free of objection would definitely be of considerable value, but the change must originate in the industry itself in view of the large commercial interests involved. All that can be attempted at present are some suggestions along this line. With the exception of Angelica, which is apparently a native name, 43 and muscatel, which is derived from a varietal name, the type names of California dessert wines are derived from European wines. Satisfactory substitutes for type names such as port and sherry are difficult to find, even though it is known that the process of manufacture and the flavor of California sherry, for example, is so different from the Spanish that the California product bears little resemblance to its prototype. The most logical development in nomenclature seems to be to restrict ordinary commercial wines to the following names: I. Red, sweet dessert wines A. Without muscat flavor 1. Full red color — California port 2. Tawny color — was called California tawny port in the pre-Prohibition era 3. Trousseau port — usually a tawny type and seldom produced since Repeal B. With muscat flavor — red muscatel II. White, sweet dessert wines A. Without a rancio or cooked flavor 1. Without muscat flavor a) Amber or yellow color, high sugar content — Angelica o) Water-white color — California white port 2. With muscat flavor — muscatel B. With a rancio or cooked or caramel flavor 1. Below 7 per cent sugar a) 0-2.5 per cent sugar — California dry sherry o) 4-7 per cent sugar — California sweet sherry c) 0-4 per cent sugar — California sherry 2. Above 3.5° Balling a) With a pink tint — California Tolcay b) With an amber color — miscellaneous, poorly defined types 43 Amerine, M. A., and A. J. Winkler. Angelica. Wines and Vines 19(9) :5, 24. 1938. 22 University of California — Experiment Station New names could then be evolved for distinctive types or high-quality wines and a trade demand created for these products. Or, if a predomi- nant varietal origin is established and the wine has a distinguishable varietal character, the varietal name might be used. This latter type of nomenclature, with the exception of muscats, is less important for dessert wines than for table wines ; although, as indicated, a Trousseau port has been made in California and marketed as such. California Port. — Port is the appellation originally used for the sweet fortified wines originating in the Douro district of Portugal. These are always red unless otherwise denoted. This port, which is mainly drunk in England, is aged in about 135-gallon oak casks from three to ten years, or more, before being sold as ruby port or as tawny port. The former has a red color while the latter has lost its pure red color with age and has assumed a brown-red hue. A limited quantity of the best wine in the best years is aged only a year or two in oak casks, is then bottled, and is called "vintage port ;" such wines must be aged many years in the bottle before reaching maturity and, because of the heavy crusts which they throw, are decanted before serving. In Portugal there is no set chemical criterion for port as far as sugar content is concerned : some ports are made almost dry with only a few per cent of reducing sugar, while some are very sweet. Typical analyses of ports are given in table 8. The variability in composition is well illustrated by these figures. Botelho, 44 however, gives the following limits for the better port wines : alcohol, 17.0 to 22.2 per cent; total acid, 0.35 to 0.59 per cent; volatile acid, 0.025 to 0.090 per cent; extract, 10.1 to 14.3 per cent; reducing sugar, 8.1 to 11.2 per cent; levulose-dextrose ratio, 1.4 to 3.3. Rigid legislation controls the districts in Portugal within which port wines may be produced. 45 "California port" is the name applied in this state to red, sweet dessert wines which have a Balling of over 5.5? They have been made here since at least gold rush days. As table 8 indicates, the present-day com- mercial California port contains 17.7 to 21.3 per cent alcohol, 0.30 to 0.67 per cent total acid, 0.026 to 0.178 per cent volatile acid, 11.1 to 14.3 per cent extract, and an average of 10.6 per cent reducing sugar. Cali- fornia port has more acid than our muscatel or Angelica but it should not have more than 0.65 since the wine is then apt to be too tart. The color of California ports varies widely. This is due to the difficul- ties encountered in producing well-colored ports from the miscellaneous varieties grown in this state and to the differences in the length and methods of fermentation and aging used. There is a well-defined trade 44 Botelho, J. C. Etudes sur le vin de Porto. Annales de Chimie Analytique 17: 49-63. 1935. 45 Serra, E., and M. Rodriques. Legisla^ao sobre vinhos. 472 p. Empresa Juridica Editora, Lisbon, Portugal. 1938. Bul. 651] Commercial Production of Dessert Wines 23 H m fe ^ 3^ as 1 1 1 rt CO "5 1 1 1 1 1 1 -**<«>•© 1 1 1 £® £~ lO i-l CO K Sq 1 1 1 OO O M O CO oo 1 1 1 a a 1 1 1 •* w w 1 1 1 33 1 1 1 CO* ""~"T oo" 1 1 1 J* 1 1 1 -I IN (D 1 1 1 bO (O .a *> ""i<" oo i>~r a 1 1 1 1 OO Ui CO 1 1 1 ffl "a .3 k. „_._._ 8-8 un UK uk 01 io e n O ■* CO t- 1-H »0 ■>* CO lO N N M >> o o o .-i © © © © © o o> w H s. § 3 *8 UK UK un t-n '3 3 03 M "O N (D r)i OO * li .-1 © © © rt o o d 1 1 1 E-t H W o> W ™ * V. 00 02 fcn 3. u 02 f5 ° cS " 3 i-i M S3 IS lO ifl » CM OS G N W N +i *3 o § a 8 N lO T(l CO i-c CO M «5 O) h-t CO >-, So t>- 00 o D "3 21 CM CM CO CO S CO Id tl oo O 3 w N N OS ~H t-. Oi N ifl ® © fe CM r^ ^H CM rt i-H CM -h i-i «l 58 A o> fe S. B tj .5 > M « -3 > .3 >- gg< 8 s s -* o o Ph T3 >> a 9 DC bO S 8 o o a a .a .a g g-a -a S CM ,.2 .si's Is; O 3 "* 3 cp « o3 in O 0> 3 ffl a5.2 rtO* ^* «*i rt CO £ a) S ~ £ "3 JS CO CO "TJ J^ S3 I bj) O " C S 9 co -, CB T. 3 SRS=! "= o* a? •-O g, >.^ >oo 3 oJ 3 Illl a ^3 -^ CO cu § ■8 S S § N tS53^ bi 1^^ 6 si .9"sO p ft to 3 X S •gco S 0>'« CD B£* B^»§ W J1 3 W -Ssa^eS CJ H .2 ^« I 03^ § . § feW^J«2^ o-d! S"3 ^o5 o3-3 »5 24 University of California — Experiment Station demand for moderately red, bulk port. Well-aged, bottled ports of the tawny-colored type were formerly fairly important in the California market, however, and may again be developed in the future. The less- sweet, lighter-colored ports may be used as appetizer wines before meals. The richer, darker ports and the well-aged tawny ports are des- sert wines, more appropriately served after dinner. Special trade names should be developed for these quality, special-purpose wines. Angelica. — The origin of the name Angelica is in doubt, but it ap- parently has been used continuously since Mission days. The type used to be the sweetest made in the state. Indeed, it originally resembled the fortified musts (mistelles) of France, Spain, and Italy, which are mainly used for blending. Since Repeal, there has been a tendency to allow the musts intended for Angelica to ferment before fortification. Table 9 shows that the present-day Angelica is not as sweet as the pre-Prohibi- tion product, indeed, not any sweeter than muscatel (table 11). If it conformed more nearly to its fortified-must character, it would be still sweeter, averaging over 13 per cent reducing sugar or approximately 8° Balling. The wines of this type have a low total acid but should not be below 0.30, and for their best flavor probably should average 0.40 per cent or higher. Their color is golden, similar to that of the muscatels. They should be free of too much aldehyde aroma or of a distinguishable sherry or rancio character. Because of the uncomplicated, clean, rich, bland character of Angelica, it should not be a difficult wine to produce well. The place of the Angelica has always been as a dessert wine. It finds some use in the winery for blending purposes. As to quality, the Angelica has little to recommend it. It is too sweet for proper aging, and it lacks any distinctive varietal character. Commercially, however, there is a demand for a very sweet white wine, which the Angelica admirably fills. California White Port. — White port is occasionally produced in Portugal, but white varieties of grapes are used and it resembles a less- sweet, aged California Angelica, having a pleasing light-amber color and frequently a slight muscat flavor. It is limited in production and importance there. California white port has been produced in California for many years for a limited market. Here, however, the wine usually has a water- white appearance. Such a color is not possible in normally produced and aged wines, and to secure it the use of decolorizing charcoal is necessary. In pre-Prohibition days animal charcoals were mainly used for the de- colorization. Since Repeal, decolorizing charcoals of vegetable origin have been used for its preparation. The objections most commonly raised by the public to such a wine are : that it does not conform to any Bul. 651] Commercial Production of Dessert Wines 25 1H lb 1 OS OS CO ^ 8^ 1 1 1 S~ M «0 a to to OS OO N 1 1 1 1 O) » N pq -^ g 5 s A S * M CD CO 00 CO o 1§ Ol H U) H * (O >> o o o -h © © « -»-i o t» 'S c Oj Tt< 3 & 8 be 3 So CO ■^ t*i 1 1 1 CO «»■* < o> o ^ v. H O c3 CO i-H •* Tt< CO CM fe (H So O) il N its © CO o a o 3^ H o3 & 8 U5 CO O CD CO 00 Q 1§ e>-« 00 y-t CO © m a eS # £ "a a 03 oo a a a 03 03 CO CO a o •2 a a 3 a a 5 3 3} JJ 3 3 i» a a a I ^ a a If liat Ph § a < 2 '3 '3 | s £ P4 ©.- W-3CO .§ -5,2 • £ 3« £3 *" PQ >,2 5 a ® ^ c Jd a ° ^33& O 3W » TJ .00 g * fe 2 - ° d&£ 8 l H I I ■e.s| & } ^«^w i c ^ <3 *l (M Eh « O C0 Ph S3 A 8 . "^ © *es co 1 3 Ig 1 1 © CO •t" oo 00 <3»-h 3 i> 43 s 02 o a u W c3 " o ft o to fc 3^3 A § m n m (D N H o Eh So "tfi £gg< 5 § s<^ t3 o o Ph 3 -S .2 & 8 3 ai co C •3 < < T3 co 03 « c3 3 .^^ a 31 § °°8 « a § 3 If as aj _coT3 »4 28 University of California — Experiment Station l - S o o CO © c5 1 1 1 1 > 1 l l >o V Ol 1 1 1 I I i 0> rn o © (-- WN N oc co co fto C3 ~-i © © co t^ CO "5 ■*f © »o CM f~ © >i -i d o o © o rH ©' O rH © © 3 C3> a "3 a o3 H 1§ «3) S © § © OS i— i U5 © © © 1 1 1 1 co 1 ° © 5 3 ^ 8 00 CM CM ■*< © Tt< <* eq a 5 " e^T as So en": •^ 00 rH -8 t~ OO © a QO 1Q CO CD t^ ©~ »; Is Ol C N N H K CO CM 0C CO t^ 1-H i-H i— i-H rH i-( CM rH ,-h 1-1 C3> o g CU io a c © T* •** t^ (M © cr " co" T* o 5 3 » 11 »-! co oc O t~ « 00 i-l US oc CM t^ CM -H r- CM i-H i— 3 0) & „ a 8 ++ ++ -H B= B= *- 15 * » « <* N - © ■* O- t^ t^ -* ft o CM CO O- ■<)i N CC CO - •f CO © x a | | "ft "ft 3 3 03 03 71 •2 "0 OS £ 1 £ £ 3* .o '43 £ 6 £ 1 "ft 9 s £ 8 I r 3 ^ 3 3 J ■3 3 3 J « 3 d a. 3 ft .3 -§ 3 3 S £ 3 3 S D 1 3 3 § D S £ 3 of '3 fi 1 :i & S 3:3 S CM •-< .3 t- - e3 C a 53 "g 3 1 o i^S< >;£ s h a •3 3 "-4 O o o c » ^•S ? l> J 03 .3 . "o 8 -o 3 .2 .2 ^ § 4 8-S8 9 co'O e-g o .2 Sh3 a v « C r] 3 Sol. O 3 0> 3§\S w o 'ad 8 P3 '^ftCJ -g.9| S tH c S© a'-' -J 33 .2=3 ■SO 03 .3 1 — ' PQ ® . 2a, Oh c?ft CO 43 "0 T3 O 3> g »s .3 S© -s" 5 8-a J 3 -2 |3KH ISl .2 3o il^ 03 a ft t^ S3 CM CO rj 00 © oTfi i« tic O «?* 8 S S ■SSJ'I -sis .-sag lSas-3 3 9 1 o- ^ 3 : 03 •< O a xi cfi O Bul. 651] Commercial Production of Dessert Wines 29 California Sherry. — The origin of this type name is obviously Spain, but the method of production utilized here more closely resembles that used in Madeira. 47 Sherry of the drier types as produced in Spain during the last century and a half has involved a maturation in contact with a film yeast (called the flor in Spain) combined with aging in a solera system. The film-yeast stage results in the production of a considerable amount of aldehydes and other flavoring constituents. The solera system (see p. 99) is really a complicated system of blending in which casks of similar types of wine but of different ages are placed in a series. Mature wine is drawn from the oldest cask, which is then filled from the next oldest, and each succeeding cask is filled from the next oldest. A good solera may contain as many as ten stages. The sweeter and more alcoholic types of Spanish sherries are also aged in a solera system for a number of years, but they do not usually pass through a film-yeast ("flowering") stage; although some wine which has undergone the film- yeast stage may be blended with the sweeter types of wines before bot- tling to give them more character. Analyses of typical Spanish wines are given in table 12. The composition of the various Spanish types is not at all uniform, particularly with respect to the alcohol content. This variability arises from the fact that fortification, if made, is carried out in unequal steps, depending on the quality of the wine, and also from the apparent fact that certain Spanish sherries increase in alcohol con- tent during storage. (See p. 15.) Wines of a similar type are produced by analogous processes in Aus- tralia and South Africa, and Schanderl 48 has indicated the general utility of the film yeasts for this purpose. Hohl and Cruess 49 have demon- strated that the process is applicable in California. Although sherries produced by film yeasts will undoubtedly be of increasing importance in California, there is no indication that the shift to this process will be very rapid. The usual baking process will probably continue to be the major method of production used in this state for some time to come. When properly made and aged, the product made by baking is a dis- tinctive type of wine. California sherries at the present time are wines of about 20 per cent alcohol which have been baked two to four months at 120° to 140° F. They should be nearly dry or only slightly sweet (not over 4 per cent reducing sugar) in taste. California sherries with below about 2^2 per 47 Castella, F. de. Madeira. Victoria Department of Agriculture Journal 26:577-87. 1928. 48 Schanderl, H. Untersuchungen iiber sogenannte Jerez-Hef en. Wein und Rebe 18:16-25. 1936. 49 Hohl, L. A., and W. V. Cruess. Observations on certain film-forming yeasts. Zentralblatt fur Bakteriologie, Parasitenkunde und Infektionskrankheiten 101: 65-78. 1939. 30 University of California — Experiment Station <& 00 Rc5 >i U t~ »-H >C III O CO OS fsl Ol H M III M - iH ^ o> © CO > t~ H t>. 111 O — IQ til ill 0) <3 t~ III -1 III III _3 ^ * ^8 e= w=- w=- CN © CO CN CN OO N CO N Ol T»< -H CO Tjt U5 § So C3 ~* ^H OO OS N CO CO CO OO o Ol rt U5 © ^h lO >> nod d o d — © — © © © —< © © 3 C& S-. a '3 5 03 H CO o « * CO CO M< Ol M «5 So o o o o o o o © © o © III III o> 03 io a 8 * O _^ CO c5 © CN t^ Ol 00 CO Tf Ol «o t~ t~ r-t CO © i-H X |8 t^ CO ■* © CN CO O0 CO "5 © — i "5 O 1) N CM CO y-< CM H t _ o ,3 o CO oo to 00 0O N CO CM © © CO ^H ■<*< co >o © o o rt OO Ol — I CO 00 CM 00 © ■* ■"*< I-- © t-H lO ON rt ON cm rt ^H CN ^H <-H C<, _ _ c<, ^ rt < » a © fc IS A S OO (0 N t~~ CO © N CO N 2o So CO CXI 1> O i-h tjh CO CM CO III III >-i o o —1 © © i-H © O III III I 03 O © *~ 3^ *8 H Ol M © 00 CO CO N H 00 CO lO «5 CO •* CM Ui •& »-l ■>* i-H OO N ^ 0O IN lO © So ON-* © % 13 2 Ph -i co O a J3 J 'a 3 03 to d "a 1 co s >> co W5 CO Ol i s £ <5 a CM c" 'a TB A "» 3 £ >> 3* _o 3 CO co C co >> E £ .2 -a ]B co r3 £ £ '■0 is £ £ E £ 03 3 3 a 3 3 p ,j=! 3 5 i a ^ 3 3 ® m 3 3 S. • S « a a S 03 o s I i^ £ s I \ 1 .aa^afi^ .. & s s i £ 3 :§ £ '3 "i 1 1 | g£ B '3 ^ f\< B '3 a 03 .-h > ^j 03 .-h > S S < g § S < o |sa< .2 & .3 u .s|ss *% g p 91 co 3 M cu IS P. 3 rt ^^ CtH ;3 cp 2 co.2^' ^ 2 ■s!!°: .Sg H 2 CD CO cu ii si 3 M 3 3 •C,;3o h3 oS- 1 ^? co co 3 3 " ■31*3 II 1 lilt i ffi ^'- 53 3 O 3 Bul. 651] Commercial Production of Dessert Wines 31 I cent reducing sugar are considered to best represent the California dry-sherry type. Sherries of 2% to 4 per cent sugar taste sweet and are commonly sold as California sherry. The proposed new standards for California sherries set the maximum sugar content for the class at 7 per cent; dry sherries to contain to 2% per cent; sweet sherries 4 to 7 per cent. As indicated in table 12, there is considerable unnecessary overlapping of composition between these types at the present time, and the limits of sugar concentration suggested above might be more uniformly utilized. The present maximum limit of 4 per cent sugar is likewise poorly enforced : a number of California sherries exceed this limit. The color of California sherries should be a pale amber. Darker wines are usually of poorer quality due to baking at too high a temperature or aging too long in small cooperage. Light-colored types can sometimes be produced by baking at a lower temperature or by the use of charcoal. The latter procedure would be undesirable even if it were permitted. Miscellaneous Wine Types. — The commercial production of the Cali- fornia Tokay, Madeira, Marsala, Malaga, and similar types has been very small in this state. Furthermore, there has been little unanimity of opinion within the industry as to what each type should represent. Certain of these types have been produced by the addition of reduced must, while others are sherry-flavored blends. If there is a demand for a sweet wine to which reduced must has been added, the development of a new name would be advisable, since the California wines made by this procedure do not closely resemble their European prototypes. The variability in composition of these types when produced in their original home is considerable (table 13), but the wines within each type retain their similarity owing to the process of production — Madeiras, owing to their baking; Marsalas, owing to the addition of reduced must; and Malagas, owing to the very sweet grapes used as well as to the addi- tion of reduced musts. The use of these type names for California wines is confusing. If wines of related characteristics are produced, they should be given a new and distinctively California name. The California Tokay is an entirely different product from the Hun- garian Tokay. This latter wine is a natural, sweet, unfortified wine largely produced from the Furmint grape. The California wine is a fortified sweet wine and is almost always a blended wine. It is not ordi- narily produced with the Flame Tokay grape, for this grape is ill-suited to the production of a sweet dessert wine because of its low sugar con- tent. Most California Tokays are a blend of Angelica, California port, and California sherry. Only enough port is used to give the wine a pink tinge. The sherry reduces the sugar content of the Angelica and gives 32 University of California — Experiment Station 6* eocooo H"0 0> "o3 PQ 1 1 1 1 1 1 1 CO—ICO 1 1 1 OOcO«S "« a cd a 8 ^ON ^foO^H OS CO * 11 1 1 1 CM CO t~- 1 1 1 O-* t~ t^wco >> -HOO i-ido odd O !- d 5 8 o3 H ScSS 1 1 1 OO ^f «5 OOO 1 1 1 S .8S d 1 1 1 1 1 1 d C3> & fe A 8 •f t~- o oor^ co M 3 CO 11 St 1 1 1 N1DO) OOCNCO 1 1 1 COO0 CO tod* 1 1 1 s» +5 03 « 8 t^ OS »0 t^-HCO MXrt CN OICO CO ^H CO CO CO O! o-*m llCOOi a> cm co •^'oo'^h "o «g rCJ g § cot>.cc OOrHH CO CO CO COt~- CO ■* » T3 v. 1 'c3 A cj !OM« COO ^ oooot^ oot^ CO CO * d I m co -«t d cOtj< «; TfN« OO CO »fl e *-< O t. H c» ?. J ft « a 1 s 1 00 'o. "ft CO G a a c3 S3 co cc •£ '£ a C? X V O '-3 _o .o 1 d 'Jo 3 1 % A 1 ^ 3, § O (-, Ph o a 03 02 a 1 3, B 1 CO CN s *aa Sflfl *aa »as 2a a «!33J CO 3 3 a oj33 J 103 31 laal fHSgO S§ fl g D °:aa^ 1 81 | - s JJ 1 3 r> M .2.1 S 3 |gi-s> 2«2 c3 '^ s Ah 'S'3'S & (_ 03 .-h > c3U 13 3 s 3 fc^ 1 O 6 3® ^ & S N P W^ O d Sj ^3 'C d 03 ^S s as >> co 2; ^ CD c3 - S ^2 5*1 % .51 23§ 03 fl .Si 8 O m d 3 CD s 1?« b 00 a b 00 bC 8 -d c 3 ^3 O C d!z;_N "ft a 03 5 mS d ®si ^0 a -^ to o3 *« fl 2 > d CS O CD 3 *o CD M 03 .ScS" CD 0^ a^ . 0) ISg^ >>^ >1 -a 5£a*^ -5 c d 'ft a >, 03 C ■* 2 8 1 a a 03 >1 jH d a> d c3 Ilia >S2 l/ i a cd . +j les on e on 4 e on 1 CM "o cu >> ^ a g ci *^ CD ■ £3 D. M M H d oj sam vera vera > ^J-d .-CD eo o3 03 c3 u 03 a w O ^ T3 TJ •Codec! o3 -_j3 O o3 03 ■£ o> o> cu to cq oi oJ rt <3 > O U COO »« © CM OS g ■*»< Oi . ^( C» l 5 ^8 t^ 00 OO * OO •>*< g§ co oo »»h r^i o 3 3 g » rt CM - © © 5» 1 o H a. u ++ ^ O IS N IS M 115 © W <--> *# •*< CO lO CO •** 1 s ° °" C3> « CO S _c 6 c 2 E P C y CO is o a > _,_ fa -3 # 1 - a t .2 ^ 3 ^ S| 13 * > II 9| 8*. I g |a 1 8 2] "* 03 t» >_ el * H ^a g co co o> I .-h OS CO-, o3 -r 03 p © co 03 I" o3 © gco co c .2 S M 4) l| CD O aa II a a © _■ S-3 as .a a 03 03 - 9 fa d -J3 J.OCO O « J2 •M- C ° g o Bul. 651] Commercial Production of Dessert Wines 35 development of varietal flavor in the grape when the fruit is fully ripe. Satisfactory varieties for dessert wines not only should be as sweet as possible, but also should not be too acid. Some varieties are too high in acid, even at advanced maturity, for producing dessert wines. (See citation in footnote 13, p. 13.) In addition, the variety should have the flavor and color requisite for the type of wine for which it is to be used. Naturally its viticultural characteristics, such as production and disease resistance, must be favorable if it is to be grown profitably. Effect of Environmental Conditions. — The desirability of a given va- riety will depend not only on its inherent character and composition TABLE 15 Suggested Composition of Musts Destined for Dessert Wines Type Balling Total acid as tartaric pH Minimum Balling acid degrees >26.0 >26.0 24.0-28.0 25.0-29.0 grams per 100 cc 0.50-0.65 .50- .65 .50- .65 0.50-0.65 pH <4.00 <4.00 <4.00 <4.00 ratio 40 40 37 38 but also on certain environmental conditions under which it is grown. Among these are the climate and the seasonal and soil conditions. The warmer districts are generally conceded to be the most desirable for grapes destined for sweet wines. In California this means the warm interior-valley area and the viticultural areas south of the Tehachapi Mountains. Even varieties which normally are used for table wines when grown in a sufficiently cool district may be utilized for dessert wines when grown in a warmer district. Some varieties of dessert-wine grapes may, however, be grown in the cooler districts and still find a satisfactory usage in dessert wines, especially in the warm seasons. This is particularly true for red dessert wines, for grapes grown in the cooler locations are richer in coloring matter than those grown in the warmer districts. 51 And some varieties may be used either for table or for dessert wine, according to the seasonal conditions and the time of picking. The varietal recommendations listed below refer only to grapes produced under conditions at least as warm as those of Lodi and Davis. In table 15 are given some suggestive limits for the composition of musts destined for dessert wines. (See also Bulletin 639, cited in footnote 4, p. 3.) Muscat -flavored Varieties. — The Muscat of Alexandria is the domi- 51 Winkler, A. J., and M. A. Amerine. What climate does — the relation of weather to the composition of grapes and wine. The Wine Eeview 5(6):9-ll; (7):9-ll 16. 1937. 36 University of California — Experiment Station nant muscat-flavored variety planted in California. Wines made from this variety have a distinctive or even a strong varietal aroma. There is not only a large supply of this variety available, but also, in general, it can be obtained in a good condition at reasonable prices. Some objection may be raised to the variety because of (a) the lack of refinement in its wine, (b) the extreme raisining which occurs in years of small crop or during a particularly hot season, and (c) the ease with which its musts spoil. The lack of refinement in aroma and flavor appears to be an in- herent quality of the variety. The extremely low total acid and very high pH of musts from this variety, notably in grapes picked late in TABLE 16 Analyses of Muscat-flavored Grape Varieties, Davis Period tested Average date of harvesting Balling Total acid pH Balling Quality of wine Variety acid years 5 5 6 3 6 6 3 4 Oct. 1 Sept. 26 Sept. 27 Sept. 26 Oct. 12 Oct. 25 Sept. 26 Oct. 9 degrees 28.5 26.2 27.6 27.8 25.5 25.5 26.7 25.8 grams per 100 cc 0.57 ,49 .66 .57 .59 .47 .38 0.49 pH 3.59 3.72 4 03 3.65 3.85 3.81 3.83 3.87 ratio 50 53.5 41.8 48.8 43.3 54.3 70.3 52.6 Fair, lacks color Good Good Muscat Hamburg Muscat of Alexandria Muscat St. Laurent Fair, lacks color Fair Fair Good the season, partially accounts for the frequent spoilage of its musts. The Muscat Canelli (Muscat Frontignan in France) is, from the flavor standpoint, more desirable than the Alexandria. It is unfortunately a small producer, ripens very early, sunburns easily, and is available only in small quantities. The wine is, however, of appreciably higher quality, being much fruitier and better balanced than that of the Muscat of Alexandria as well as being of a more distinctive flavor. Malvasia bianca is also a promising muscat-flavored variety. Its wine is distinctively flavored and the vines are fair producers. Moreover, it does not sunburn easily, although it becomes very high in sugar. The wine of the Orange Muscat is also distinctively flavored and the vines produce well. The Muscat Hamburg and Aleatico are the chief red grapes with a muscat flavor planted in California. The wines of these varieties usually are too light in color to be used by themselves and must be blended with deeper-colored types. Of the two, the Aleatico has a more refined aroma and flavor. Table 16 gives some comparative analyses of the various muscat- flavored varieties. Varieties for White, Sweet Dessert Wines. — The chief types of white, Bul. 651] Commercial Production op Dessert Wines 37 sweet dessert wines are Angelica and the California white port. The characteristic of these wines is that they have a large residual sugar con- tent. Grapes destined for these wines must therefore have a high sugar concentration, as is indicated in table 15. Thompson Seedless (Sultanina), Mission, and Grenache (when it is fermented off the skins) , are the most important varieties ordinarily util- ized for these wines in California. The best-quality wine appears to be that of the Mission, the product having a fruity flavor and a desirable smooth texture. The musts from Grenache are sometimes too colored TABLE 17 Analyses of Grape Varieties Commonly Used for Angelica, Davis Variety Period tested Average date of harvesting Balling Total acid as tartaric pH Balling acid Quality of wine Black Prince* years 3 5 5 3 6 6 2 5 Sept. 19 Oct. 7 Sept. 28 Oct. 3 Sept. 25 Sept. 10 Oct. 9 Sept. 16 degrees 25.1 23.3 24.4 26.8 25.6 23.7 26.3 24.9 grams per 100 cc 0.43 .67 .45 .55 .38 .52 .54 0.67 pH 4.22 3.42 3.77 3 46 3.91 2.95 3.72 3.72 ratio 58.3 35 54.3 48.7 67.3 45.6 48.7 37.2 Erbalus di Caluso Grenacheft Fruity, tart Good Grille- Thompson Seedless Verdelho Fair Good * Davis, Lodi, and Fresno averaged. t Skin pigmented but white musts are obtainable by careful harvesting and rapid crushing and pressing. j Davis and Fresno averaged. for use in a white wine. The Thompson Seedless is used mainly because of its availability but is not regarded with favor by the industry. When properly matured, it produces merely a neutral-flavored must and wine. Other varieties that are not commonly used in California but which attain a high degree of sugar include : Erbalus di Caluso, Sauvignon vert, Grillo, and Verdelho. Analysis of varieties which may be used for this type of wine are given in table 17. Varieties for White, Dry Dessert Wines. — California sherry and re- lated wines of less than 6 per cent sugar are included in the white, dry, dessert wines, commonly used as appetizer wines. As indicated in table 15, these types of wines do not require as high a sugar content in the must as do other types of dessert wines. Varieties which are naturally not so high in sugar, as well as the usual varieties when picked somewhat earlier, may be used. In any case, however, the Balling-acid ratio should not be below 35 nor should the total acidity exceed 0.60 or 0.65 at the most. Wines with a higher extract content may be made by using musts of high sugar content, if the fermentation is controlled so that it does not stick before the sugar content is sufficiently reduced. 38 University of California — Experiment Station Rocques reports 52 the Palomino to be the most important variety in the sherry district of Spain, with only small quantities of the Mantuo de Pilo, Mantuo Castellano, and Perruno. Some Pedro Ximenes is planted but the chief locale for this variety in Spain is in the Malaga district. The most important varieties which are used for these types of wines in California are Palomino, Thompson Seedless, Mission (when it is fer- mented off the skins), Feher Szagos, Flame Tokay, and Malaga. The last two varieties are much less satisfactory, since they are ordinarily TABLE 18 Analyses of Grape Varieties Commonly Used for White, Dry Dessert Wines, Davis Variety Boal di Madeira Burger Feher Szagos* Flame Tokay t Green Hungarian . . Inzolia bianca Malaga Malmsey Mission! Palomino Thompson Seedless Period tested Average date of ripening Oct. 1 Sept. 21 Sept. 5 Oct. 7 Sept. 29 Oct. 4 Oct. 25 Oct. 17 Sept. 25 Oct. 4 Oct. 9 Balling degrees 23.8 22.0 22.1 19.6 19.6 23.9 20.0 24.4 25.6 24.0 26.3 Total acid grains per 100 cc 0.58 .39 .45 .40 .49 .46 .49 .38 .46 0.54 pH pH 3.78 3.59 3.42 3.54 3.91 3.80 3.77 3.91 4.13 3.72 Balling acid ratio 41.1 32.2 56.5 43.5 49.0 48.7 43.5 51.5 67.3 52.2 48.7 Quality of wine Good Poor, thin Fair, if grapes clean Poor Fair Good Poor Good Good Very good Fair * For Fresno. t Davis and Lodi averaged. t Davis and Fresno averaged. below the minimum sugar requirements (table 18). The Feher Szagos rots badly in many vineyards, and in such cases it, too, is undesirable. Muscat wines have occasionally been used for California sherries. The muscat flavor partially remains even after baking. For this reason, their use in regular sherries should be avoided. But the development of a distinctively flavored sherry containing Muscat wine, properly named, would very possibly be desirable. Varieties for Red, Sweet Dessert Wines. — California port is the only common red dessert wine produced in this state. It requires musts of high sugar content, but the grapes should be free of overripe or raisin flavors. In the pre-Prohibition era, Trousseau was considered to be the best variety. Its wines are fruity and smooth but usually are deficient in 52 Eocques, X. Les vins de liqueur d'Espagne. Eevue de Viticulture 19:446-53. 1903. Bul. 651] Commercial Production of Dessert Wines 39 color. This difficulty was formerly overcome by marketing the aged Trousseau wine as tawny-colored port. Post-Prohibition requirements have been so largely for highly colored ports that blending has usually been resorted to in order to increase the color, and the desirable qualities of the Trousseau wine have thus been masked. The varieties commonly used to increase the color have been the Ali- cante Bouschet, Alicante Ganzin, Grand noir, and Salvador. Only the first and last of these four varieties are available in large quantities. The Alicante Bouschet is undesirable because of its flavor and its tendency to deposit its color during aging. The Salvador is satisfactory for blend- ing when used in minimum amounts. It is somewhat darker than the Alicante Bouschet and has a slight fruity, aromatic flavor which may be undesirable in ports. Other varieties commonly used are the Carignane, Petite Sirah, and Zinfandel. The Carignane makes only an ordinary-quality dessert wine. The Petite Sirah is useful if the grapes do not sunburn too much. The Zinfandel produces a desirable, distinctively flavored, fruity wine, but unfortunately, the condition of Zinfandel grapes produced in the warm interior districts is usually unsatisfactory, particularly with respect to bunch rot. Other varieties not commonly used but which have desirable qualities for this type of wine are the Tinta Madeira and Valdepenas. Muscats should not be used in California port. The development of red muscatels, however, is a satisfactory innovation, particularly the fortified Aleatico which has a pleasing fruity flavor. The composition of typical varieties used for the production of California port is given in table 19. According to de Castella, 53 the most important varieties for port in Portugal are Alvarelhao, Bastardo, Touriga, and Tinta Cao. Only very limited amounts of any of these varieties have been planted in Cali- fornia with the possible exception of the Bastardo, which Olmo 54 believes to be very similar to or the same as the Trousseau. A test of these varieties may reveal some to be of value in California. VINDICATION Dessert wines may be made from partly or wholly fermented musts or from unfermented grape juice. The extent to which they are fermented varies with the initial Balling degree of the must and the type of wine to be made. The fortification of practically unfermented free-run must drawn off the pomace does not yield wines of as good a flavor as those produced by the fortification of slightty fermented wines. 53 Castella, F. de. Notes on Portuguese varieties of grapes. Victoria Department of Agriculture Journal 14:398-408, 565-70, 622-28, 673-86, 731-40. 1916. 54 Olmo, H. P. Le Trousseau et le Bastardo. Eevue de Viticulture 87:174-75. 1937. 40 University of California — Experiment Station Possible Use of Concentrate. — Sweet wines which derive their sweet- ness from the unf ermented sugar remaining in the wine after the fer- mentation has been checked by the addition of fortifying brandy are superior in flavor to those prepared by the addition of grape concentrate to a dry wine. It has been claimed that the latter wines are fuller- bodied and more fruity than those made by the normal process. This is not true in general, although the use of concentrate for the amelioration TABLE 19 Analyses of Grape Varieties Commonly Used for California Port, Davis Variety Period tested Average date of ripening Balling Total acid tartaric pH Balling acid Color in- tensity* Quality of wine Alicante Bouschet Alicante Ganzin.. Cinsaut Carignane Grand noirf Mourisco preto . . . Pagadebito Petite Sirah Petite Bouschet . . Tinta Madeira Trousseau Salvador} Valdepefias Zinfandel years Oct. 8 Oct. 7 Sept. 28 Oct. Oct. Oct. Sept. 29 Sept. 20 Oct. 4 Oct. 14 Sept. 30 Aug. 28 Sept. 26 Sept. 19 degrees 21.6 24.1 25.3 23.4 23.3 23.8 22.4 27.2 25.2 25.9 29.1 24.5 24.3 23.1 grams per 100 cc 0.60 .71 .49 .53 .58 .42 .73 .49 .57 .56 .79 .55 0.70 pH 3.74 3.54 3.76 3.71 3.78 3.97 4.04 3.61 3.73 3.81 4.09 3.70 3.87 3.45 36.2 33.9 51.6 37.2 44.0 41.0 53.5 37.3 51.4 45.5 52.5 31.2 44.2 33.0 457 558 130 211 277 175 461 637 421 206 165 1,430 191 Poor Fair for color Good but lacks color Fair in warm years Poor Good but lacks color Fair, little flavor Too tart and sunburns Fair Good, distinct Good, lacks color For color only Good Fair if grapes are clean * These figures represent the relative intensity of color expressed on an arbitrary scale: the higher the figure the greater is the concentration of pigment, t Davis and Manteca. t Delano figures only of the wine when intelligently practiced is not harmful and may even be desirable for certain types of wine. Although Federal Regulations No. 7 permit the use of "pure boiled or condensed grape must or pure crystallized cane or beet sugar or pure dextrose sugar, at the time of, or after fermentation," 55 the sugar must not be in excess of 11 per cent of the weight of the wine. This restriction on amount of sweetening does not apply to grape products. California regulations prohibit the use of any sweetening agent other than grape concentrate. Sweetening agents, if any are used, may be added only prior to fortification, but see page 136. 55 United States Bureau of Internal Eevenue. Kegulations No. 7, relative to the production, fortification, tax payment, etc., of wine. 188 p. (See especially p. 39 and 91.) United States Government Printing Office, Washington, D. C. 1937. Bul. 651] Commercial Production of Dessert Wines 41 Period of Fermentation. — The period of fermentation for most sweet dessert wines, whether white or red, is usually short; hence the musts are fermented on the skins save where very light color is desired. Where light color is desired or for sherries, the crushed grapes are allowed to stand only until the cap forms and rises to the surface, and the free-run must is drawn off to be fermented separately. Fermentation on the skin yields wines containing more of the fruity flavor of the berry and is necessary for color extraction in the case of ports. To dissolve flavor and color, time is required. Since the must is usually fortified soon after active fermentation sets in, there is only a restricted period during which the juice and skins are in contact. The lower the initial Balling degree of the must and the warmer the grapes, the shorter is this fermentation period. To extend the period of contact, the grapes should be crushed when as cold as possible, and 4 to 6 ounces of potassium metabisulfite (or its equivalent of sulfur dioxide) added per ton of grapes directly after crushing. Good contact between the free-run juice and the skins must be established, and the fermentation should be conducted with a submerged cap or the cap punched down frequently and well. The crushing should be conducted so as to completely crush all berries and break the skins, without, however, breaking the seeds. Special procedures to insure adequate color extraction from red-grape skins are often used in preparing port-type wines. Control of Fermentation. — The same principles as are necessary for making dry table wines of quality apply to the production of good dessert wines. Selection of grapes as to kind, quality, and maturity, fer- mentation with the utmost care and cleanliness at cool temperatures and with selected strains of yeast, and proper care during aging is just as necessary for one type as for the other. "Wines of quality are made; they do not just happen," to quote Humboldt. 56 The prevailing highly competitive dessert-wine market unfortunately has resulted in placing greater stress on the more economic production of sound wines for the cheaper trade than on production of wines of quality. Both the sound or- dinary dessert wines available at a comparatively low price and the fes- tive wines are needed, but a better balance in the amount of the two types produced in this state is necessary. Too little attention is paid to the conditions of fermentation of dessert wines, particularly to the fermenting temperature, which often becomes so high as to favor development of harmful bacteria. When grapes are crushed at high temperatures, the must may stick in fermentation owing to the rapid development of acetic acid bacteria, which quickly form sufficient acetic acid to check growth of yeast (see p. 143) . At these high 56 Humboldt, E. Wine making is an art. Chemical and Metallurgical Engineering 43:651-53.1936. 42 University of California — Experiment Station temperatures (90-100° F), the chemical activity of the yeast is also altered, so that abnormal flavor develops. To avoid undesirable effects of microorganisms, the grapes should be crushed at as low a temperature as possible and the must cooled as directed for table wines. (See Bulletin 639, p. 48-52.) Dessert-wine musts, however, need not be cooled to as low a temperature as those for table wines, fermentation temperatures below 85° F being satisfactory. Use of Sulfur Dioxide. — Sulfur dioxide should also be used to help control the activity of undesirable microorganisms. 57 The amount of sulfur dioxide added should be the minimum necessary to inhibit bac- terial growth. Usually 4 ounces of potassium metabisulfite per ton of grapes (2 ounces of liquid sulfur dioxide) is sufficient and the amount should not exceed 6 ounces. The must should not contain too much sulfur dioxide at the time of fortification as otherwise the resulting wine will have a bad flavor. In the case of muscatels, Cruess (cited in footnote 15, p. 00) recommends the addition of 100 parts per million of sulfur dioxide (7 ounces of potassium metabisulfite per ton) com- mercially, although as little as 50 parts per million (3 ounces of metabisulfite) prevented bacterial development in wines fermented ex- perimentally at temperatures of 85 to 90° F. The amount of sulfur dioxide necessary is greater the warmer the grapes and the higher their Balling degree. After early fall rains larger amounts may also be required. Lightly fermented musts tolerate less residual sulfur dioxide than more completely fermented ones. A small amount of residual sulfur dioxide is not harmful even in the distilling material, for it combines with the sugars and aldehydes present in the wine and improves its keeping quality. Excessive amounts in the distill- ing material, however, corrode the plates and interior walls of the still, yield unpleasant-smelling volatile sulfur derivatives (mercaptans), and result in fortifying brandies that do not blend well with the musts. The extent to which the addition of 4 ounces of potassium metabisul- fite per ton of grapes reduced the volatile acidity of Fresno sweet wines during the vintage season of 1936 is shown in table 20. The average (cal- culated as the mode) for the sulfited wines was 0.0349, that for unsul- fited 0.0399. There were too few samples of the latter, however, to obtain a good comparison. Cool fermentation and the use of sound grapes com- bined to yield the favorable results reported for the unsulfited wines. These results indicate that under sanitary conditions sulfur dioxide need not be used regularly in the fermentation of musts for dessert wines. 57 Cruess, W. V. Observations of '36 season on volatile acid formation in Muscat fermentation. Fruit Products Journal 16:198-200, 215, 219. 1937. Theron, C. J., and C. J. G. Niehaus. Wine making. Union of South Africa Dept. of Agr. Bui. 191:1-98. (See p. 58 and 61.) 1938. Bul. 651] Commercial Production of Dessert Wines 43 During the vintage season of 1934 and 1935, much higher volatile acidi- ties were found in the freshly fortified, unsulfited wines. Use of Pure Yeasts. — Pure wine yeast is not widely used at present in fermenting dessert wine musts. The grapes for these wines are picked at a stage of maturity when the numbers of yeast cells present on the grapes are high and no difficulty is experienced in starting the must to ferment. A few wineries use starters to obtain faster fermentations so that the f ermenters may be reused more frequently ; or to obtain wines TABLE 20 Influence op Sulfur Dioxide on the Volatile-Acid Content of the Finished Wine in 12 Fresno Wineries Treatment and winery no. Samples Volatile-acid content of wine Range Average With metabisulfite* No. 1 number 9 4 9 9 17 5 12 6 6 77 4 7 5 16 per cent 0.009-0 048 .021- .033 .004- .066 .019- .066 .013- .048 .020- .041 .020- .090 .023- .040 .030- .068 .004- .090 .016- .031 .024- .059 .039- .108 0.016-0.108 per cent 0.032 No. 2 .027 No. 3 .032 No. 4 .045 No. 5 .031 No. 6 .033 No. 7 .041 No. 8 .034 No. 9 .047 Total .035 (mode) Without metabisulfite: No. 10 .026 No. 11 .041 No. 12 .073 Total 0.040 (mode) 4 ounces of potassium metabisulfite per ton of grapes used. that clear better owing to the use of pure wine yeasts that agglomerate or clump together and settle out as a granular deposit rather than yield a finely dispersed cloud of yeast cells. The use of strains of yeast selected so as to develop the most flavor from the available variety of grape is still in its infancy. The use of such selected strains of yeast might help to improve the flavor of the available dessert wines. 58 Fermentation By-Products. — The short fermentation period, the con- version of only a part of the sugar initially present, and the dilution during fortification combine to reduce the concentration of the by- products of alcoholic fermentation in dessert wines. The partially fer- mented musts are lower in volatile acid, fixed acid, and glycerin content than are the completely fermented wines. A significant change also ^Joslyn, M. A. Biological control of fermentation. Wines and Vines 19(4): 16-17. 1938. 44 University of California — Experiment Station occurs in the type of reducing sugars present. The levulose-dextrose ratio increases from slightly under 1 to much more than 1, and this increase is reflected in the change in degree and even sign of the optical rotation of the wine (which becomes more levorotatory; see p. 16). By-products of fermentation which accumulate at a faster rate in the early stages of fermentation influence the flavor of dessert wines much more than those which accumulate in the later stages of fermentation. Yeast strains forming aromatic flavorful constituents in the early stages of fermentation will therefore be most useful. Influence of Acidity. — It is commonly believed that the acid content of the musts for dessert wines should be low since a high acid content does not yield a well-balanced young wine. If a low-acid wine is matured for several years, however, it becomes flat to the palate and lacks f ruiti- ness. High acidity is also useful in suppressing the growth of harmful bacteria. Theron and Niehaus (cited in footnote 57, p. 42) recommend an acidity of 0.60 per cent as tartaric for dessert wines. Fornachon 59 recommends the addition of tartaric acid to must to insure that the pri- mary fermentation will take place under conditions favorable for the optimum flavor production with the minimum of autolysis of yeast cells. A wine made from a properly balanced must is clearer, more stable, and more resistant to bacterial attack. Cruess (cited in footnote 15, p. 14) found that the addition of citric acid produced muscat wines of lower volatile acidity. Fermentation. — As a rule the crushed grapes are fermented on the skins for 2 to 3 days, the free-run wine is then drawn off (approximately 140 gallons per ton of grapes) . The pomace is not pressed to obtain wine because such press wine would be too harsh in flavor, and of poor qual- ity. The remaining pomace is covered with water and allowed to ferment dry with occasional stirring; this requires another 3 days, and the low- alcohol wine produced is used exclusively for distilling material. The free-run pomace wine is drawn off and the pomace washed or pressed (see p. 56). Distillery Capacity. — The capacity of the distillery determines the quantity of wine to be made, for there should be an ample supply of fortifying brandy on hand to fortify at the right time; or, if a sufficient amount cannot be produced, arrangements for the purchase of the needed amounts should be made. Twight™ considers an adequate supply of fortifying brandy to be the most important factor in the production of good dessert wines and in the economical operation of the winery. He 59 Fornachon, J. C. M. Bacterial fermentation in fortified wines. 19 p. Australian Wine Board Publication on Wine Investigations. 1938. (Mimeo.) eo Twight, Edmund H. Sweet wine making in California. Part II. California Grape Grower 15(11) :4-5, 7. 1934. Bul. 651] Commercial Production of Dessert Wines 45 estimates that a winery having a distillery with a capacity of 2,000 gal- lons of 90 per cent (180° proof) fortifying brandy a day should not handle over 100 tons of wine grapes a day if maximum production and quality is desired. If too large a tonnage is crushed, there will be insuffi- cient brandy to stop the fermentation at the desired Balling and the wines will be below standard in sugar content. There is also danger that the wine will be allowed to ferment until it gets "stuck," and when such wine is fortified it yields a product of poorer aroma, flavor, and keeping quality. Use of the Hydrometer. — As in the making of dry table wines, the course of fermentation is followed by the Balling or the Brix hydrometer and a long-stem or indicating thermometer. After the grapes are crushed and before fermentation sets in, the approximate sugar content of a representative sample of the must is determined by the Balling hydrom- eter. In the absence of alcohol, the Balling degree is higher than the actual sugar content owing to the presence of acids, salts, and other soluble nonsugar constituents. The Balling degree often rises shortly after crushing, especially if raisins are present, owing to extraction of sugar from the nonjuicy, high-sugar grapes. With the onset of full fer- mentation, the Balling degree drops, not only because of decrease in sugar content but also because of the formation of alcohol, which lowers the specific gravity of the solution. In a partially or completely fermented must, the Balling degree is lower than the actual sugar content. The extent of such reduction varies with the relative alcohol and sugar content; 61 at a given sugar concen- tration, the higher the alcohol content the lower will be the indication of the Balling hydrometer. In spite of this discrepancy, the Balling hydrometer is commonly used as an indication of the extent of fermenta- tion in both dessert-wine and table-wine making. The Balling degree of a wine or must may be corrected for the effect of alcohol by the procedure suggested in paragraph 329 of Regulations No. 7 (cited in footnote 55, p. 40). This is based on the assumption that the effect of alcohol in decreasing the specific gravity of water solution is independent of other solutes, that is, that the specific gravity of a solution of sugar and alcohol is lowered by alcohol to the same extent as is that of water alone. The specific gravity of the solution then varies from that of water by the difference between the increase caused by the presence of sugar and the decrease caused by the presence of alcohol. It is assumed that not only are the effects of the individual constituents additive, but also that volume changes during solution are negligible. The relation used is as follows : 61 Joslyn, M. A. Approximate alcohol tables for wine. California Grape Grower 15(11) :12-13. 1934. 46 University of California — Experiment Station Specific gravity of dealcoholized wine = specific gravity of wine - spe- cific gravity of an alcohol solution containing the same percentage of alcohol as the wine + 1. To use this relation, it is necessary to have tables showing the specific gravity of solutions of alcohol and of cane sugar at various concentra- tions. (See tables 33 and 35, p. 154 and 158, in this publication, and tables 4 and 5 in Regulations No. 7.) Three examples will illustrate this calculation. Example 1. Suppose that a wine tests 6° Balling at 60° F and contains 12 per cent alcohol. The specific gravity of such a wine is 1.02320; the specific gravity of the alcohol solution of equivalent alcoholic strength is 0.98430. Subtracting the latter from the former leaves 0.03890, and adding 1 gives 1.03890, the specific gravity of dealcoholized wine. This corresponds to 9.9° Balling. Example 2. Suppose that a wine tests 6° Balling at 60° F and contains 21 per cent alcohol. The specific gravity of such a wine is 1.02320; the specific gravity of the alcohol solution of equivalent alcoholic strength is 0.97496. The specific gravity of dealcoholized wine is then : 1.02320- 0.97496 + 1, or 1.04824, which corresponds to 12.2° Balling. Example 3. A partially fermented must for port tests 14.1° Balling and contains 6.4 per cent alcohol before fortification. The specific grav- ity of the dealcoholized wine would then be 1.0560-0.99098 + 1, or 1.06502, which corresponds to 16.3° Balling. FORTIFICATION Fortification is the process of raising the alcoholic content of a par- tially fermented wine or of an unf ermented must by addition of fortify- ing brandy to a concentration sufficiently high to check the fermentation and to prevent ref ermentation. Legal Restrictions. — The quantity and kind of brandy used in fortify- ing wine is restricted by federal and state agencies. For the purpose of taxation, three classifications of wines are established by United States Internal Revenue law: wines containing not more than 14 per cent alcohol by volume, wines containing more than 14 per cent and not exceeding 21 per cent, and wines containing more than 21 per cent and not exceeding 24 per cent. Under the Definitions and Standards for Wines of the California Department of Public Health, the commercially recognized, named types of dessert and appetizer wines are required to have a range in alcohol content of 19.5 to 21.0 per cent by volume (±0.5). Altar wines produced for ecclesiastical purposes may contain as little as 18 per cent of alcohol. A few states have slightly different re- strictions on the maximum and minimum alcohol contents of dessert Bul. 651] Commercial Production of Dessert Wines 47 wines and, if shipments into such states are contemplated, details should be obtained directly from the proper state authorities. According to Theron and Niehaus (cited in footnote 57, p. 42), South African law requires fortification to at least 17 per cent of alcohol. Amount of Alcohol Required. — The concentration of alcohol neces- sary to prevent the fermentation of fortified sweet wines depends on the strain, physiological condition, and numbers of yeast cells present; the composition of the wine, particularly its acidity, pH value, and nitrogen content; and temperature and other external environmental factors. Yeasts are less sensitive to alcohol than many bacteria and most other fungi ; still most yeast fermentations are strongly retarded at 5 to 6 per cent alcohol and practically cease at 14 per cent. Alcohol-tolerant wine yeasts, however, are known, particularly the Brazilian Logos yeast, and these produce up to 18 per cent alcohol by direct fermentation of must containing sufficient sugar (30 per cent or over). Even the common wine yeasts can produce, under special conditions, wines of 20 per cent alcohol content. 62 The strongly respiratory aerobic yeasts such as the apiculate yeasts and Saaz-type beer yeasts, on the other hand, are in- hibited at alcoholic concentrations of only 4 to 6 per cent. Theron and Niehaus recommend that wine be fortified to at least 17 per cent of alcohol by volume to prevent fermentation. This concentra- tion may not be high enough to check the fermentation, particularly with active wine yeast, which may grow in wines up to 18 to 20 per cent alcohol. 63 Ventre 6 * points out that the abrupt addition of 15 per cent alcohol to a medium which is not in fermentation inhibits subsequent development of yeast; but when the addition is made to a medium in ac- tive fermentation the process is at first arrested and then continues at a slower rate until the maximum quantity of alcohol tolerated by the par- ticular strains of wine yeast, approximately 16-18 per cent, is formed. Euler and Lindner 65 point out that the sensitivity of yeast to alcohol 62 Cruess, W. V., E. M. Brown, and F. Flossfeder. Sweet wines of high alcohol content without fortification. Journal of Industrial and Engineering Chemistry 8:1124-26.1916. Hohl, Leonora, and W. V. Cruess. Effect of temperature, variety of juice, and method of increasing sugar content on maximum alcohol production by Saccharo- myces ellipsoideus. Food Research 1:405-11. 1936. Hohl, Leonora. Further observations on production of alcohol by Saccharomyces ellipsoideus in syruped fermentations. Food Eesearch 3:453-65. 1938. 63 Kohl, F. G. Die Hefepilze, 343 p. (See especially p. 150 and 152-53.) Verlag von Quelle und Meyer, Leipzig, Germany. 1908. 84 Ventre, Jules, Traite de vinification pratique et rationnelle. vol. 1. Le raisin. Les vinifications. 490 p. (See especially p. 108-9.) Librairie Coulet, Montpellier France. 1930. fi5 Euler, Hans, and Paul Lindner. Chemie der Hefe und der alkoholischen Garung. 350 p. (See p. 283-84.) Akademische Verlagsgesellschaft m.b.h., Leipzig, Germany. 1915. 48 University of California — Experiment Station increases with increase in sugar content of the medium. Hartelius 68 indi- cates that acids as well as alcohol inhibit growth and fermentation. Both Hartelius and Rahn 87 indicate that the gradual retardation and ulti- mate cessation of the rate of fermentation before all the fermentable substrate is used up is due to the toxic action of fermentation products, particularly of the alcohol. Growth of yeast ceases at a lower concentra- tion of alcohol than does the fermentation produced by the yeast, al- though the rate of fermentation gradually decreases as the concentration of alcohol, whether added or formed, increases. The danger of spoilage of sweet dessert wines by growth of lactic acid bacteria, which grow in wines of low acidity even up to 20 per cent alcohol, has prevented the gradual raising of the alcoholic content by the periodic addition of small quantities of fortifying brandy, as is practiced in European dessert-wine production. Gradual fortification is possible there, because of the higher acidity of these wines and because the smaller cooperage used there facilitates handling the wines at lower temperatures; although cases of spoilage of fortified wine are not un- common in these countries. When fortifying brandy is added in install- ments so that fermentation continues at a diminished rate for a week or over, the brandy blends better with the sweet wine. The product is mellower and softer than one of the same alcoholic content obtained by adding all the alcohol in one application at the end of the process. How- ever, autolysis of yeast occurring during the retarded fermentation ob- tained by gradual fortification may contribute to the spoilage by lactic acid bacteria, according to Fornachon (cited in footnote 59, p. 44). Furthermore, Hohl and Cruess experienced difficulty in preventing "mousiness" of wines of high alcoholic content obtained by siruped fer- mentation : that is, where grape concentrate was periodically added dur- ing the fermentation. Gradual fortification is also more expensive than is a single, abrupt fortification. The California sweet, dessert wines are fortified to an alcoholic con- tent of 20 to 21 per cent by volume; occasionally lots may be fortified to as low as 19 per cent or as high as 24 per cent for blending purposes. The standard wines are marketed at 20 per cent alcohol, California port testing 6° Balling, muscatel and Angelica 7° Balling or over, and sherry 1° to -3° Balling. The higher the alcohol content obtained by the fortifi- cation, the more costly is the process, and the longer the time required to age. Wines to be consumed fairly young are more pleasing to taste at 18 per cent than at 21 per cent alcohol. 60 Hartelius, V. The growth of yeast in synthetic media and the factors produced by yeast which limit this growth. Carlsberg Laboratoire Comptes-Eendus des Travaux 20(7): 1-43. 1934. 67 Eahn, Otto. The decreasing rate of fermentation. Journal of Bacteriology 18:207-26. 1929. Bul. 651] Commercial Production of Dessert Wines 49 The quality of the fortifying brandy has a great effect on the flavor of the dessert wines. Only sound clean brandy free from objectionable flavors should be used and the quantity added must be at a minimum to reduce dilution of the wine flavors. Fruity-flavored brandies obtained from the varieties of grapes used in making the particular wine are pre- ir icr 8° r .fe f 3° 16 , « h 1°— -4--" I 6 „ £- £*~ r l?° t £^ I ^ bfi si co o ■* CN oo t» CO CD os t^ CO CO d 2 05 > 3 ° sss — i o o CN CN CN © © CN CN d d CN CN d o CN • 00 t^. t^ CO > CO | | 60 «< -<3 -3 PQ # 13 CO 00 ■* «o ^ CN t-H OS t^ 5 d 03 2 1-H t~- 1-H O 1 1 OS OS oo CN U5 £ 1 ■a 1 2oo 1 ~~ CO oo r~-i O CO CN CN T "* T T 1 o bO CO c3 co os t^ i« — i 00 o O os eo »o £ s~ co * m S ' ~ CM t>- 00 lO «o d b0 .5 > 1 *"* 03 PQ .2 N M (O © CO ■** OS OS O co lO ^-1 '2 M d S P4 V O CO «5 CN CN H © if CO d oo d t>i cn eo CN .-H i— i g> N tO CO © P-. t^ Itl 22 ice 4i d cn cn t i-i O0 if t^ «n CO ^H 1 1 (V d V ^1 f N N HUJIN O CN CO 1« CO OS d 8 "o A o o ^ >C OO U5 ■* 115 IN r-l t^ lO Ir^ "3 «0 « p bo 11 —i CN CO CO «C Tf rt CC ■* "5 t~- rH 3 d 03 3 u "o Co a 5. o> o o N COCO d CD 00 «5 lO t~ .15 ti H © CO CO OS O d cn d CN d ■* d 3 "3, a to 1 in s oo o "* *H N CO N N IO CO ^ c<; CN CO -H WJ 03 QQ e >1 <-, © d > o 11 CO c CO c 2 § c d 7 1 7 OS o- O ©" O OS §§ CD -H OC co2cl co 0< OS 1 « oc OS 1 i-l OO d o 1 eo O I eo » oo ff O CV 1 cc 1 CO .. O0 OS cs w o- os o oo o- £ t-h CO y- i-l CO i-i CO '- +3 CO " . OS » OS .. OS h OJ . "o o ft >> 5n ""' C 5s *"' C ..fee o3 rt C S H o •• d T3 a b^ E ti i g g d i ft . rt - d ft 'S cS & co M fa J!,3£ m 03 r >- 3 CC ft J h3 fa Ph CQ S < & 56 University of California — Experiment Station fications at different periods of the month, are forwarded to the nearest laboratory of the Alcohol Tax Unit of the Bureau of Internal Revenue (Federal Building, Civic Center, San Francisco) for testing. The alco- holic content as determined later is usually somewhat lower than that obtained by the gauger because of inaccuracies in the rapid method used by the gauger. The wine after fortification is released for transfer to the storage cellar. There are many variations in the details of fortification : the type and size of fortifying tank; the method of adding the fortifying brandy, whether first or last; the method of mixing; and the period of time the wine is allowed to remain in the fortifiers. Several of the pre-Prohibition wine makers preferred to run the fortifying brandy into the tank first and then add the wine and allow the mixing to occur naturally by diffu- sion. Their wines were also frequently stored for longer periods of time in the fortifying tanks before transfer to the storage cellar. Settling. — During fortification there is a rapid flocculation of yeast, proteins, and gums; and it is preferable to allow the fortified wine to settle in the fortifying tanks for at least 24 hours before it is pumped out, to allow these insoluble matters to deposit. This period is sufficient to bring to a close the fermentation, which continues for a time even in the presence of the high concentrations of alcohol. According to most modern wine makers, the sooner the newly fortified wine is separated from the crude lees, after the lees settle, the better is its flavor and the more rapidly it mellows during storage. Where the fortifying room is small or is in constant use, the wine is pumped out of the fortification tanks into storage immediately, and is given its first racking in the cellar as soon as possible after settling. Production of Distilling Material. — The distilling material, from which the fortifying brandy is obtained, is usually prepared from low- alcohol wine made from the pomace left in the fermenters after the free-run wine has been drawn off for fortification. Water is added to the f ermenter in quantity sufficient to cover the pomace, and the pomace is well mixed with it by thorough punching or pumping over. The mix- ture is allowed to ferment out dry. To obtain the maximum yield of alcohol, it is essential that the sugar in the pomace be completely fermented out. When the fermentation is completed, the free-run pomace wine is drawn off and the residual al- cohol in the pomace is removed as completely as practicable. This may be done in several ways. In one procedure the pomace is pressed out in a continuous expeller press and then washed in a counter-current type of washer (diffusion battery) to remove the residual alcohol, which may run as high as 5 per cent. In another, the pomace, after the first refer- Bul. 651] Commercial Production of Dessert Wines 57 mentation, is mixed again with water, allowed to ferment for an addi- tional 24 hours, and then pressed. The extent to which alcohol recovery may be practiced is determined by the tanks available for pomace wash- ing and by the time and labor available for this purpose. In some cases the fermented pomace is ground or otherwise disintegrated, is allowed to ferment for about 24 hours more, and is finally run directly into a pomace still. The seeds are crushed by this procedure. The fermentation of the distilling material should be carefully con- trolled so that it will not be subjected to bacterial decomposition by either the acetic acid or lactic acid bacteria. Sour or "mousy" distilling material yields a fortifying brandy of objectionable flavor that should not be used in making dessert wines of high quality. Small quantities of sulfur dioxide may be used in controlling bacterial development, but it is best not to use sulfur dioxide or to use it only for sprinkling over exposed pomace. Temperature control and fermenting with submerged pomace are preferable to the use of sulfur dioxide for obtaining sound distilling material. It is generally necessary to crush additional grapes to produce the brandy needed for fortification, and cheaper table grapes or packing- house culls are used for that purpose. These should be sound and, to expedite fermentation, they should be mixed with an equal volume of water. The lower the initial sugar content, the more rapidly will the must ferment out dry and the less will be the chances for contamination because of reduction in time of contact. Prolonged fermentation of dis- tilling material is uneconomical and undesirable for other reasons. The distilling material is adjusted to an alcoholic content of about 6 to 8 per cent, according to the operating characteristics of the still, and the alcohol is distilled off in a continuous single-column still at a proof varying with the type of distilling material and the requirements of the wine maker. Quality of Fortifying Brandy. — The poorer the quality of the dis- tilling material available, the higher should be the proof of the fortify- ing brandy obtained in order to avoid contamination with objectionable volatile acids, higher alcohols, aldehydes, and esters. The highest proof obtainable by direct distillation is approximately 190° (at 68° F). Such a spirit is very neutral in taste and is free from all congenerics, such as aldehydes, esters, and fusel oils. If the distilling material is sweet- smelling and free from objectionable flavors, a fortifying brandy of 175° to 180° may be made. This will have a higher concentration of congenerics. Fortifying brandy of alcoholic concentration lower than 175° proof is not ordinarily used because it may dilute the wine too much and give a product of thin body, particularly when the original 58 University of California — Experiment Station wine is low in nonsugar solids. Aldehydes and fusel oils are objectionable in a fortifying brandy, particularly when present in high concentra- tions, because such a brandy will not quickly blend well with the wine. Some wine makers, however, prefer fortifying brandy of 175° to that of 180°. (See p. 78.) The characteristics (such as proof, congenerics) of the best fortifying brandy to be used for a given wine are not known. Experience has shown, however, that better-flavored dessert wines are obtained when they are fortified with fortifying brandy made from the same variety of grapes as is used in making the wine. This is particularly true of fruity wines such as muscatel. A neutral-flavored fortifying brandy, however, blends better with most wines than one that is either off-flavored or that contains too high a concentration of esters, such as may be ob- tained from an aromatic grape like the muscat varieties. The longer the brandies are aged, the better will they blend with the wine. Aging of fortifying brandy, however, is seldom practical. AGING The process of aging of dessert wines, like that of table wine, consists in the elimination of undesirable substances — such as suspended yeast cells and other organic matter, excess cream of tartar, tannin — loss of harsh yeasty and new brandy flavors, and the formation of aromatic compounds particularly agreeable to the taste and smell. The rate of aging of the dessert wines is slower than that of table wines and the conditions under which they are aged are quite different. Settling. — The newly fortified wine is racked off the lees that settle out after fortification and is pumped into storage. If the storage space in the fortifying room is limited, fining and filtration may be used to produce a more rapid flocculation of the crude lees. The prompt removal of the yeast cells from the new wine is desirable to prevent autolysis (or endogenous metabolism) of the yeast cells. Autolysis liberates strongly reducing enzymes and proteins, which not only inhibit oxidation of the wines and produce unpleasantly tasting and smelling compounds, but also favor the growth of lactic acid bacteria and so render the wine more susceptible to bacterial spoilage. After the wine is pumped into storage, it is allowed to remain in full tanks for about 4 to 6 weeks and is then racked from the lees. The tanks should be kept filled as far as possible, just as for unfortified wines, to avoid the excessive oxidation that occurs in the presence of too much oxygen and to minimize the increase in volatile acidity, which, although it occurs at a much lower rate than is the case with table wines of lower alcoholic strength, does take place. After the first racking the wines in Bul. 651] Commercial Production of Dessert Wines 59 the various tanks are classified by tasting and analyzing, blended, fined with gelatin and tannin or bentonite, and then stored. Treatments during Aging. — During aging dessert wines are given special treatment required to bring out the characteristics desired. The flavor of the California-sherry-type wines is developed by a process of caramelization and oxidation at moderately high temperatures. Heating is also used to develop the flavor of Marsala and Madeira types but should not be used for the Angelica or port types, which are best if they age naturally. If certain colloids are present, pasteurization may be prac- ticed to coagulate them. It is also sometimes used to bring about a better blending of the fortifying brandy and wine, but in any case it should not be prolonged. The slow oxidation of wine and the low rate of esterification under the ordinary storage conditions result in a slow natural aging. The size and kind of tank used for storage and the storage temperature are important factors in determining the rate of aging. It is well known that wine mellows more rapidly in small tanks than in large, possibly because of the more ready access of oxygen by diffusion. Containers. — Oak is the preferred of all woods, and oak ovals are prized even today as storage containers for wine. It is believed that the oak exerts a beneficial effect on the wine. Although the presence of certain oxidizing enzymes in wood was shown to aid the aging of sake 70 (a Japanese fermented cereal beverage), it is improbable that similar enzymes are present in the old and much used oak ovals. Furthermore, particular care is taken to remove from new containers as much as practical of the oak extractives, which would give the wine an unpleas- ant harsh flavor. Storage in oak even for a short period improves the flavor of dessert wine; the quantity of oak extractives that may be present in a wine without injuring its flavor varies with the type. The heated wines with a rancio flavor tolerate more oak flavor than fruity wines like muscatel and Angelica. Chemical Changes. — The conditions of aging and the changes that result in aging of dessert wines are similar in general to those obtained in table wines. (See Bulletin 639, p. 89-91). During aging, decreases occur in extract, alkalinity of the water-soluble ash, total tartaric acid, cream of tartar, color, and tannin. These decreases are all accounted for 70 Higasi, Tuneto. The oxidizing action of "Sugi" wood. Institute of Physical and Chemical Eesearch, Japan, Bul. 8:831-38. 1929. Abstracted in: Chemical Abstracts 24:458. 1930. Yamada, Masakazu. The oxidizing action of cryptomeria wood. Agricultural Chemical Society of Japan Jour. 6:168-77. 1930. Also in: Agricultural Chemical Society of Japan Bul. 6:9-11. 1930. Abstracted in: Chemical Abstracts 25:718. 1931. 60 University of California — Experiment Station by oxidation and the precipitation of wine lees composed of cream of tartar and oxidized color and tannin. Volatile acids and volatile esters increase slowly. Heating increases the formation of volatile neutral esters. The high alcohol content of the dessert wines favors the forma- tion of larger amounts of esters than in table wines. (See p. 19.) Alde- hydes and acetals form more rapidly and in higher concentration than in table wines. The changes in alcohol content depend on the initial concentration, size of container, thickness and nature of the walls of the container, and the humidity and temperature of the surrounding air. Dessert wines stored in large containers in atmospheres of low relative humidity decrease in alcoholic content, but when stored in small oak barrels they may increase in alcohol content. The limiting factors are the differential rates of diffusion of alcohol and water through the pores of the container walls and the relative rates of evaporation at the outer surface. 71 The pores of white oak offer more resistance to passage of alcohol than of water because of the higher surface tension and viscosity of alcohol ; and at the surface the rate of evaporation of water will depend on the relative humidity. 72 (See also p. 15.) Time. — The amount of aging required depends on a number of fac- tors. The first consideration is the quality of the wine. Poor wines do not adequately profit by aging and should be blended and finished for sale early. Good dessert wines increase in quality for a number of years, the rate of aging depending on the factors previously mentioned : type and size of container, methods of handling, temperature of storage, humidity, and type of wine. Where the container is very large, dessert wines may be kept a num- ber of years with relatively little aging. On the other hand, when the wines are heated, aging is rapid but the quality is not always increased. (See p. 61.) Wines should not be aged in wood so long that they acquire undesirably high woody aromas or flavors. Foaming. — Newly fermented wines contain surface-active substances which lower the surface tension of the wine and cause it to foam during agitation. This may cause inconvenience in the handling of the wine in the winery. On aging, however, this tendency to foam is decreased, and well-aged wines do not froth or foam. Even champagnes, although they retain carbon dioxide gas like beer, do not foam as the gas is released, unlike beer. Quick Aging of Wine. — Filtration or fining or both to remove sus- 71 Fabian, F. W. The role of wood in the ageing of wine. Wooden Barrel 3:14, 22-23, 25, 27, 29. 1935. 72 Henrichs, C. G. No substitute for white oak barrels for aging whiskey. Bever- age News 33(3) :14. 1934. Bul. 651] Commercial Production of Dessert Wines 61 pended material more rapidly, refrigeration to remove excess cream of tartar more rapidly, aeration to hasten oxidation, and pasteurization have all been used in hastening maturation of wine. Only a few of the processes suggested for quick aging are desirable or practicable. Per- mission from the Alcohol Tax Unit of the Bureau of Internal Revenue must be obtained before applying any procedure. Heating 73 in the pres- ence of air and oak extractives may be used for some sweet wines, par- ticularly those like sherry in which caramelization is not objectionable. But because of the danger of forming excess acetaldehyde, and of dis- turbing the delicate balance between the constituents of the wine, which would result in subsequent clouding, quick aging is not recommended for the better wines. 74 All of the quick-aging processes now in use are defective in that they merely increase the rate of the oxidative changes and do not materially increase the rate of the particular esterification processes which give the beverage its delicate bouquet and aroma. Thus the quickly aged beverages may be palatable, but they lack the quality and character of the naturally aged product. There is some evidence, however, that the esterification processes may be speeded up by the use of catalysts. More promising are the possible applications of biologically controlled aging through the use of selected strains of various microorganisms. Of the various electrical processes tested in the laboratory, the most promising at present is an electrolytic process in which an attempt is made to combine the desirable features of the hydrogen reduction proc- ess with that of the oxygenation process. 75 The wine is reduced at one electrode and oxidized at another. This process brings about an especially rapid blending of the fortifying brandy with the wine. Other electrical processes including ozonization have been suggested, but for the most part they are not suitable to California conditions. WINERY DESIGN, EQUIPMENT, AND OPERATION 76 The general principles governing the location, size, and construction of wineries for the production of dessert wines are similar to those for plants producing table wines, which have been previously discussed in Bulletin 639, with the exceptions noted on the following pages. 73 Cruess, W. V., and M. A. Joslyn. A method to age wine. Fruit Products Jour- nal 13:365, 379. 1934. 74 Fain, J. Mitchell, and Foster Dee Snell. Artificial aging of spirits. Industrial and Engineering Chemistry, news edition 12:120-22. 1934. Joslyn, M. A. Possibilities and limitations of the artificial aging of wines. Fruit Products Journal 13:208, 241. 1934. 75 Joslyn, M. A. Electrolytic production of rancio flavor in sherries. Industrial and Engineering Chemistry, industrial edition 30:568-77. 1938. 76 General references on this subject in addition to those given in specific foot- notes in the section are listed on p. 172. 62 University of California — Experiment Station DESIGN The layout of a winery producing dessert wines differs considerably from that of a plant devoted to table wines. The generally larger size of such wineries and the greater degree of departmentalization which is necessary are primary differences. Prevailing Hind h rl/n/oao'/np p/otform Mile yafia cruster-ste/nmer c Petf yrope crvsner- stem/ner || (Pomace \cowaj/or || \\ || ||_ . 0/ stance reau, by I/ne'er writers JGauccbs Office 1 i I I H"" _ Room \ I J I &*** I I ■ Dcpovr | fi00M | Sue* A* 'Baking " PooM m m Brandy Storage Poom 8/iANOY SrORAOC PoOM Fig. 3. — Schematic floor plan of a winery and distillery for the production of dessert wine, vermouth, and brandy. A diagrammatic sketch of a plant handling from 7,000 to 10,000 tons of grapes per season is given in figure 3. The fermenting room in the sketch contains twenty 8,000-gallon open concrete fermenting vats and four closed tanks of the same or larger size. The cost of concrete f erment- ers is somewhat greater than that of redwood f ermenters. But if a suffi- cient number of concrete f ermenters are constructed, the cost per tank is reduced. Further, their cost of upkeep is less than that of redwood and their utilization of space is better. If smaller fermenters are desired or if special types of fermenters are constructed, they can be constructed Bul. 651] Commercial Production of Dessert Wines 63 at the end of the fermenting room nearest the presses, and two or all of the f ermenters at that end of the room may be removed. Production of about one-half million gallons of dessert wines a year, during a season of not less than 8 to 10 weeks, is contemplated. In the storage section of the winery, concrete, redwood, and oak stor- age containers are shown. The total capacity of this portion of the cellar is about one million gallons. This size of winery presupposes the sale of about one third of the new wine in the year following its production Fig. 4. — Aerial view of a typical large winery and distillery. and of about 15 per cent in each year from the second to the fifth year. The remainder is aged for special bottlings or for blending purposes. About 15 per cent each year is either lost in racking or distilled because of defects. Such a winery would dispose of its ordinary wines in bulk during the first year and would gradually build up a stock of quality dessert wines. The rate of turnover of wine in many California dessert wineries is far too rapid for producing quality wines. More space for barrel and puncheon storage should therefore be provided. Other departments common in such wineries are still, fortifying, brandy-receiving, machinery, boiler, and sherry rooms. Space is also provided for a cooper shop (in the storage room), a vermouth depart- ment, a shipping and bottling room, and for office, sales, and laboratory space. The sketch in figure 3 is diagrammatic and is intended only to indicate the general relation of the parts of a winery to each other. The actual layout of a somewhat larger commercial winery is shown in figures 4 and 5. For the production of exclusively bulk-quality wines, a much simpli- fied form of design is possible. The size of the container is increased — 64 University of California — Experiment Station more and larger concrete f ermenters and storage containers being pro- vided — and the puncheon and barrel rooms are removed. No adequate cost surveys are yet available to indicate the overhead of the very large winery as compared to the medium-sized or small plant. The general tendency for wineries producing dessert wines has been towards plants exceeding one million gallons in storage capacity. For plants devoted entirely to the production of bulk-quality dessert wines, table 25 is suggestive of the type of containers required. The larger the winery the greater the amount and proportion of very Tot a/ storage 2,600,000 £<*/. farms/) ting room 700,000 ya/. pump mfsey[}\^u Fig. 5. — Floor plan of the winery shown in figure 4 ; the number and size of the cooperage in the various cellars is indicated. large containers required. Plants exceeding one million gallons in capacity will find it convenient to have a number of very large con- tainers available for blending and storage. In the design of a new winery, the services of a competent architect are very desirable. The particular site and the peculiar needs of the proprietor can then be satisfied. No matter what the present intentions of the owner, provision should always be made for future expansion. The majority of present California wineries are the result of remodel- ing previous wineries or buildings. Many are therefore not economically arranged nor properly constructed. Modernization of these plants would be an advantage, reducing the cost of production, increasing the quality of the wines, and facilitating the operation of the plant. Bul. 651' Commercial Production op Dessert Wines 65 EQUIPMENT Crushers, Stemmers, and Fermenters. — The equipment for a dessert- wine plant is fairly simple, but it must be ample, both in amount and capacity. The crushing and must-pump facilities particularly should be adequate to handle peak loads. Usually it is desirable to provide separate crushers for red and white grapes. Two moderate-sized crushers are better than one very large crusher. The must lines from the crusher- stemmer to the fermenting room should be of corrosion-resistant metals which can be easily cleaned. Adequate fermenting-room facilities (see fig. 6) are not only of great assistance to the wine maker, but make possible the more intelligent TABLE 25 Number and Size of Containers for Bulk Wineries of Various Sizes 100,000-gallon winery 500,000-gallon winery 1,000,000-gallon winery Tanks Size Tanks Size Tanks Size number 1 6 3 gallons 25,000 10,000 5,000 number 2 10 2 6 gallons 40,000 35,000 25,000 10,000 number 15 9 2 3 gallons 40,000 35,000 25,000 10,000 Source of data: Levine, S. Efficient winery operation. The Wine Review 6(5) :7-9. 1938. handling of the musts so that better wines are produced and there is a better utilization of the pomace material. The size of the fermenters should ordinarily not exceed 10,000 gallons and, for special fermenta- tions, fermenters of only 500 to 1,000 gallons are useful. Sumps adequate in number and size should be provided for drawing off wine and for transfer of wine. Special diffusion-tank systems for leaching the pomace have been used in some California wineries. 77 It is usually better to provide cooling coils in the fermenters, but portable heat exchangers are also useful. Special sumps with a large number of cooling coils are sometimes provided to facilitate cooling. Most California wineries either do not press or else use all of their press wine for distillation. The screw-type continuous press (fig. 7) is the best choice if the press wine is to be used for distillation. Pure-Culture Tanks. — Near or in the fermenting room, small con- tainers for preparing pure cultures should be constructed. A small concrete or glass-lined tank on top of a larger one is the usual arrange- 77 Turbosky, M. Musts for sweet wine production. The Wine Eeview 5(9) :12-14. 1937. C>6 University of California — Experiment Station Fig. 6. — Fermenting room with large concrete fermenters. Note particularly the drainage sumps under the fermenters. This arrangement saves considerable space. (Photograph by courtesy of the Wine Institute.) Fig. 7. — Two large continuous presses. The press at the right has the cover removed to show the screw. Notice also the pressed pomace coming out. The pomace drops into the conveyor located below the presses. (Photograph by courtesy of the Wine Institute.) Bul. 651] Commercial Production op Dessert Wines 67 ment. Piping from the larger pure-culture tank to the fermenters throughout the cellar is also useful. In large wineries intermediate tanks of a size sufficient to provide regularly 2 to 3 per cent culture for the fermenters are necessary. Some dessert-wine plants use their pure- yeast culture only early in the season. Conveyors. — For handling the pomace, continuous-chain conveyors running in concrete troughs are now almost universally used. Such conveyors should be kept clean during the vintage season or they will Fig. 8. — Concrete storage tanks in a California winery. Note the measuring tubes on the outside of the tanks and the gutters beneath the outlets for drainage purposes. (Photograph by courtesy of the Wine Institute.) serve as breeding places for bacteria and fruit flies. These conveyors lead directly to the presses. Small motor-powered elevators are now commonly used to raise the pomace from the floor of the fermenter to the conveyor. In some wineries the pomace is washed out of a manhole in the side of the fermenters and into a conveyor running at their base. The fermenting room should be equipped with sufficient permanently installed, corrosion-resistant pipe and pump facilities to move the fermented wine quickly to the fortifying room or to the still room. Storage Equipment. — The size of containers for the storage of the wines is particularly important in California, where redwood and concrete containers of 25,000- to 100,000-gallon capacity are commonly used for dessert wines (fig. 8). The rate of penetration of oxygen through thick concrete containers, for example, is considerably reduced. 68 University op California — Experiment Station In addition, the extraction of desirable materials from the wood is a factor which is entirely absent when concrete is used. Finally, the ratio of surface to volume is much smaller with large-sized containers, and even if the rate of penetration of oxygen is the same, less will be absorbed per gallon of wine. The amount of material dissolved from the wood is also reduced by this factor. The number of square inches of surface exposed per gallon of wine in various types and sized con- tainers is given in table 26. Vintage dates on labels will not mean the same as far as maturation of the wine is concerned when different sizes of storage containers are TABLE 26 ►Square Inches of Surface Exposed per. Gallon of Wine in Containers of Various Capacities and Shapes Type container Dimensions Capacity Inside surface per gallon Barrel 33 X 68.5 X 80 inches* gallons 50.0! 65. 8f 134. 3f 10,000 10,000 100,000 100,000 square inches 54.6 Port pipe 40 X 73 X 91.5 inches* 49 3 Sherry butt 48 X 86 X 112 inches* 38.6 112.5 X 162.5 inchest 9.32 (11.16 feet) 3 §.. .. 10.48 19 feet 4J^ inches X 29 feet 6H inches*. . (27.73 feet) 3 §. . . 4.56 4 86 * Height X circumference at head X circumference at center, outside (inside dimensions used in calculation). t Capacity obtained by gauging. J Inside height X inside diameter. § Inside dimensions. used. With wines stored in 135- to 160-gallon puncheons, three or four years are considered to be about minimum for the wine to become smooth and to acquire bottle ripeness; much longer periods are frequently used in European countries for dessert wines. Wines stored for the same time in larger containers usually are not so smooth and have an undesir- able aroma of the fortifying brandy because of their reduced rate of aging. It is not intended to imply that the rate of aging of wines is directly proportional to the square inches of container surface exposed per gallon of wine. But this and the influence of the wood itself on the aging are factors which are far too frequently neglected in the handling of quality dessert wines in California. With ordinary wines, of course, the aging is not so important and these factors can be and commonly are overlooked. The best wines should accordingly be aged in oak barrels or puncheons. The lesser-quality wines are stored in redwood or concrete containers. New Containers. — New concrete vats and tanks should be washed, soaked with a tartaric acid solution, and, if possible, filled for some time Bul. 651] Commercial Production of Dessert Wines 69 with good distilling material or preferably with clean but ordinary dessert wine. Various coatings are used by some wineries. Unless the wine is to be stored for some period in concrete, this is unnecessary. Even then it is difficult to find a permanent, nonflaking coating. Glass- lined concrete containers have not, so far, been much used in California, although they are preferable. New redwood tanks should be soaked with a warm alkaline solution Fig. 9. — Refrigeration equipment in a large California winery. Smaller Avineries can obtain smaller and more compact units for their needs. (Photo- graph by courtesy of the Wine Institute.) and water. The storage or fermentation of distilling material and des- sert wine in such tanks is also useful. Oak containers are usually soaked with an alkaline solution, steamed, filled with water, and, if possible, temporarily conditioned with clean distilling material and sound wine of low quality. Refrigeration. — The generally warm conditions of California win- eries slow down the rate of deposition of tartrates. Cooling the wines to a temperature of below 25° F hastens removal of tartrates and also usually facilitates clarification. (See p. 126.) The solubility of air in wines at low temperatures also increases the oxygen content of the wine, and consequently the rate of oxidation when the wine is once more placed at ordinary temperatures. 70 University of California — Experiment Station Some California wineries provide a room which may be held at low temperatures, while others insulate concrete tanks and equip them with cooling coils (see fig. 9). A cold room, although expensive, is a great convenience and should be provided if possible. Other Equipment. — Heat exchangers are standard equipment for modern wineries. Portable heat exchangers are useful, or a permanent installation can be arranged in the storage room and the wine pumped to it. Flash pasteurizers in which the wine is rapidly raised to a high Fig. 10. — Bottling room in a large California winery. Note the automatic bottlers and cappers. (Photograph by courtesy of the Wine Institute.) temperature and then very quickly reduced are considered most de- sirable. Numerous types of niters are found in California wineries. Plate and frame filters, in which the use of filter aids is reduced, are usually satisfactory. At least one pad filter for polishing purposes is necessary. Heat exchangers and filters should be constructed of corrosion-resistant metals. (See Bulletin 639, p. 58-59.) Rubber hoses are very important pieces of equipment. Tasteless, odor- less hoses of various sizes and lengths should be provided. Sloping racks on which the hoses can be drained after use are essential. The life of the hose can be considerably increased by proper handling, especially when not in use. Bul. 651] Commercial Production of Dessert Wines 71 Automatic bottle washers (fig. 10) are great timesavers, and their efficiency exceeds that of hand-washing. The filling equipment should be kept clean. Use of corrosion-resistant metal for the fillers is essential. SANITATION AND MAINTENANCE Winery Design. — Elimination throughout the winery of all nooks and corners which are difficult to clean is the first requirement in sani- tation. Construction of concrete piers for the vats and tanks, adequate sloping of the floors to the drains, and concrete floors are very desirable. The floors should not be so smooth as to be slippery nor so rough as to catch excessive amounts of dirt and organisms. Services. — Hot and cold water and steam outlets throughout the fermenting and storage rooms should be arranged at the time of con- struction of the winery. Regular use of plenty of water on the floors should be the standard practice. All must and wine pipes must be easily drainable and so constructed that there are no nonflushable sections. Care of Containers. — Empty concrete containers are the simplest to keep clean. A thorough washing and sterilizing with a dilute hypo- chlorite solution is about all that is necessary. Open redwood fermenters may be filled with limewater and, if the surface film of calcium carbonate is not broken, they will remain "sweet" for several months. Painting redwood fermenters with lime and allow- ing them to dry out is also a common practice. The hoops should be tightened soon after the fermenters are emptied; and the lime solution should not be too thick else it will be too difficult to remove. Closed redwood tanks are usually left empty without treatment, but they should be scrubbed, washed, and sterilized with a dilute hypo- chlorite solution before such storage. Releasing sulphur dioxide gas into the tank occasionally and tightening the hoops regularly is also helpful in keeping such containers in a good condition. Empty sherry-cooking tanks are very difficult to keep clean. Usually the measures mentioned for redwood tanks are used. Keeping these tanks full of wine is a better procedure. After a few years' use, sherry- cooking tanks leak so badly (fig. 11) that the soft staves must be re- moved and the internal and external surfaces thoroughly cleaned. Keeping empty oak cooperage in good condition is a difficult problem. Cleaning thoroughly and storing in a room that is not too dry, too warm, nor subject to drafts is the best procedure. A dilute solution of sulfur dioxide and sulfuric acid is sometimes used to fill such containers. Most wineries handling dessert wines find it possible to fill most of their empty oak cooperage with wine from concrete containers, which reduces the storage problem of empty wooden cooperage. 72 University of California — Experiment Station The services of a skilled cooper to care for cooperage is good economy in any winery. When barrels are returned to the winery, they should always be soaked with a dilute alkaline solution (such as soda ash) and thoroughly washed. If the barrel has become moldy, it may require steam, hypo- chlorite, or scraping and recoopering. Tank cars have become a source of much contamination, since they Fig. 11. — Outside of a redwood sherry tank during "cooking." Note the leakage from between the staves. (Photograph by courtesy of the Wine Institute.) are usually returned in a very insanitary condition. Distributors should be asked to clean such tanks when they are emptied, or at least to wash them out. When they are returned to the winery, they should be scrubbed, using soda ash, and the lining should be examined for any defects. 78 Returned empty containers and recently emptied storage vats, be- cause they contain alcohol vapors, constitute a fire hazard. Smoking or lighting of matches in their vicinity should be prohibited. 78 Quaccia, L. The care of cooperage. The Wine Keview 7(7) :7-8. 1939. Bul. 651] Commercial Production of Dessert Wines 73 Many dessert wineries are troubled with the growth of fungi on the outside of the tanks. Recently proprietary germicidal paints have been developed for the control of such growths in the food industries, and they should prove useful in wineries as well. 79 The cask borer (Scobicia declives) is only occasionally found in red- wood tanks. Oak tanks painted first with alum and then linseed oil are immune to attack. Keeping oak casks in the dark is also helpful in preventing attack. Antiseptics. — Cleaning solutions of varying degrees of strength are available for use in the winery. For mild cleaning, dilute solutions are satisfactory. But where extreme contamination has developed, stronger solutions are necessary. At present, for general winery-cleaning pur- poses, solutions yielding free chlorine are used in most wineries. Cleaning bottles (see p. 141) and other special problems may require a different type of solution. The antiseptic solutions must, of course, not come in contact with wines. Health of the Workers. — According to Delaunay, 80 winery workers are not subject to any particular occupational diseases. The humidity, darkness, and lack of aeration are health hazards in some wineries. A survey of California wineries by Russell, Ingram, and Dakan 81 showed that 76 per cent of the employees were exposed to alcohol vapors; 53 per cent to carbon dioxide; 21 per cent to excessive dampness; and smaller percentages to sulfur dioxide, alkalies, carbon monoxide, organic acids, and various other substances. They recommend adequate ventilation, good lighting, protective clothing, safety directors, and other measures to safeguard the health of winery workers. DIRECTIONS FOR MAKING RED DESSERT WINES 82 The chief type of red, sweet dessert wine produced in this state is California port, although a red muscatel is occasionally made. In the eastern United States, where native species of Vitis or hybrids of native species and V. vinifera varieties are grown, red dessert wines, such as Concord, Ives Seedling, and the like, are occasionally made. All of these wines resemble each other in their color, alcohol percentage of 18 or 21 per cent, and sugar concentration of at least 6 per cent. (See p. 22-24.) 79 Epstein, S. S., and F. D. Snell. Antiseptic and germicidal paints. Industrial and Engineering Chemistry, industrial edition 33(3) :398-401. 1941. 80 Delaunay, H. Hygiene des ouvriers du vin. Chimie et Industrie, Special no., p. 615-18. 1925. 81 Russell, J. P., F. R. Ingram, and E. W. Dakan. Industrial hygiene survey of California wineries. California Dept. of Public Health, Industrial Hygiene Serv- ice Investigation Eept. No. 2. 36 p. 1939. 82 General references on this subject in addition to those given in specific foot- notes in the section are listed on p. 172-73. 74 University of California — Experiment Station Varieties suitable for California port are given on page 38 and for red muscatel on page 36. HARVESTING AND FERMENTATION Grapes intended for red dessert wines should be picked and trans- ported to the winery with the same general care as those used for red table wines. Picking only good-quality fruit of suitable varieties at the proper stage of maturity — 25° to 29° Balling (see p. 35) — and trans- porting the grapes to the winery without delay or metal contamination and in as cool a condition as possible are the elementary requirements for harvesting and delivery of the fruit to the winery. Delay in delivery, so that fermentation starts in bruised fruit before crushing, is very undesirable with the fruit of low acid and high pH used for dessert wines, since bacteria develop particularly well under these conditions in the warm interior valleys of California. Furthermore, grapes for des- sert wines are harvested at a more advanced maturity, hence they are more susceptible to handling injuries than those for table wines. They should therefore be handled at least as carefully as the grapes intended for table wines. The belief that dessert wines are free of spoilage because of the short fermentation and that the grapes may therefore be handled in an insanitary fashion is entirely erroneous. Contamination of the must by improper handling frequently results in serious spoilage losses, or in wines which require time and costly treatments to finish. Crushing. — The stemming and crushing is done with the usual equip- ment and the must then pumped to the fermenters — either redwood or concrete commonly being used. Usually these are open vats. A better manipulation of the cap is possible if they are relatively shallow so that the depth of fermenting mass is small or if submerged cap fer- menters are used. A small amount of sulfur dioxide is introduced into the must while filling (see p. 42). The amount may be as small as 50 parts per million for the best grapes, since the fermentation period of dessert wines is short. But larger amounts are used under very warm conditions or with fruit of poor quality. Either potassium metabisulfite, sulfur dioxide gas, or a 6 per cent solution made from one of these may be used. (See also Bulletin 639, p. 43-45.) Pure yeast cultures (see p. 43) should be used, at least at the start of the season. They are added an hour or more after the sulfur dioxide has been applied. Two or 3 per cent of the pure yeast culture should be added. Various strains of wine yeast are used for red sweet wine, but in view of the restricted extent and period of the fermentation, there does not appear to be any advantage in using a special strain for these BUT, 651] COMMERCIAL PRODUCTION OF DESSERT WlNES 75 wines. The contribution of mixed yeast cultures to dessert-wine pro- duction is not known. If such cultures greatly delay the fermentation, their influence may be beneficial in permitting longer contact of the skins and juice and hence greater extraction of color and flavor. The possibility of flavor contamination and the formation of undesirable hazes in the wines must be taken into consideration when unrestricted wild yeast growth is permitted. CONDUCT OF THE FERMENTATION The primary problem in making red dessert wines is the adequate extraction of color from the skins during the restricted period of fer- mentation before fortification. In the making of red muscatel, flavor extraction from the skins is also important. Temperature. — The period of fermentation cannot ordinarily be ex- tended too long because of the consequent increase in the number of fermenters which will be required. Nevertheless, increasing the period of contact of the skins with the fermenting juice is a practical means of extracting a larger amount of color. A moderate temperature in the fermenter is a partial means of prolonging the period of the fermenta- tion as well as securing a cleaner fermentation. A number of methods for securing or maintaing a lower temperature may be utilized. Crush- ing the grapes in the coolest possible condition is one. Cooling either by coils in the tanks or by pumping the fermenting juice through a heat exchanger is the most satisfactory measure. The use of sulfur di- oxide alone is not a satisfactory means of slowing down the fermenta- tion, although it will have some effect through its control of the temper- ature. Overheating in the cap is to be studiously avoided, for it may result in wines of a "cooked" flavor. Sticking is not a problem in making red dessert wines. METHODS FOR COLOR AND FLAVOR EXTRACTION Management of the Cap. — Proper manipulation of the cap is a most important means of increasing the extraction of color and flavor. The traditional Portuguese method is to keep the cap constantly punched down by treading. A more sanitary substitute for this method is to use long poles which have a flat board attached to the end for punching down. The punching-down must start within a few hours after crushing and should be kept up as continuously as possible. The submerging of the skins not only increases the color extraction by increasing the con- tact of the liquid and the skin, but also increases the extraction of other soluble constituents such as sugar and flavoring matters. The constant movement of the skins through the liquid also promotes this dissolution 76 University of California — Experiment Station by the mechanical breakdown of the skins. Too much stress cannot be laid on the necessity for continuous punching down if the best results are to be obtained. This is particularly important if partially dried berries are present. A mechanical means of punching down would be very desirable, since it is very difficult to punch down the cap in a large tank. Note: Adequate safety measures to prevent asphyxiation — smother- ing due to excess carbon dioxide — should always be taken where men are working on the top or edge of ferment ers. Usually two men should work together. Modern air conditioning, such as is found in breweries, will undoubtedly be used in wineries in the future. Submerged Cap. — Although the submerged-cap system of fermenta- tion does not have the mechanical advantage afforded by punching- down, it does give a maximum submersion of skins in the juice. Tanks with removable wooden tops and concrete tanks with fixed tops through which the juice alone rises are in common use in this state. Difficulties in manipulation and sanitation of the wooden lattice have restricted their use, and closed concrete tanks are now favored. There is a raised ledge around the top of the tank and manholes in the top. A small wooden lattice may be placed below the manhole. This is easier to use than the large lattice over the whole upper surface of the container. (See Bulletin 639, p. 70-72.) Pumping Over. — Pumping over is another method used for securing color extraction. The inlet hose should be handled by a man in order to spray it over the full surface of the tank. Although pumping over aerates the must, this is not objectionable with red dessert wines unless it ex- cessively speeds up yeast growth and consequently the rate of fermenta- tion. Cooling the must while pumping it over will help to prevent an undue rise in temperature. Special Methods. — In the large winery, sufficient attention cannot always be given to the laborious details of the techniques previously described for securing maximum color and flavor extraction. In certain districts and seasons, the grapes available are deficient in coloring matter, and in such cases, special methods may be needed even in the small winery. Furthermore, sometimes deeper color may be wanted than can be obtained by ordinary methods. The following procedures are useful for this purpose, but their effect on the quality of the resulting wines is by no means clear. Such methods should, therefore, be used with caution and only after trial has established their effect on the quality, as well as on the color of the resulting wines. One of the procedures used commercially involves drawing the juice from the skins, heating it to 150° to 180° F, and pumping it back onto Bul. 651] Commercial Production of Dessert Wines 77 the skins. The color of the skins so treated very rapidly diffuses into the juice, and the whole mass can be pressed in a few hours. The re- sulting red-colored juice is handled as if it were a white must, except that it should be cooled to about 70° before adding the pure yeast culture. The heat treatment gives a fair pasteurization, and little or no sulfur dioxide need be used. In this heat treatment, a number of variables must be controlled to secure the best results. The temperature of the whole mass, after pump- ing back the hot juice, should be at least 140° F. The heating of the juice should not be done with live steam or by any system whereby a portion of the juice is overheated or burnt. Where equipment is available, heating of the entire mass of grape- skins and juice may be done in steam-jacketed vessels made of stainless steel or some other noncorrodible metal. The amount of color extracted can be increased by raising the temperature above 160° F or by in- creasing the period on the skins. Tressler, Joslyn, and Marsh 83 state that the heating should not be for too long a period nor at too high a temper- ature lest too much tannin be extracted. Amerine and De Mattei 84 have shown that color extraction can also be obtained before crushing by dipping the whole grapes in boiling water for 1 minute. The grapes are then crushed and may be pressed almost immediately, if desired. The general principle involved in all procedures in which heat is used is that of killing the cells of the skins of the grapes which contain the anthocyanin pigments so that the cells become permeable or at least lose their characteristic semipermeability, which ordinarily prevents the rapid dissolution of the pigments from the skin. In addition to preliminary color extractions, heating before fermenta- tion and not cooling the must and even purposely "sticking" are used. 85 PRESSING In Portugal, partial fortification of the whole fermenting mass before pressing is permitted. This may have the advantage of securing greater color and flavor extraction, but some of the added spirits will be lost with the pomace at the time of pressing. Under present government regula- tions, such a procedure is not feasible in California, and the must is pressed before fortification. Where ordinary-quality grapes are being used, only the free-run 83 Tressler, D. K., M. A. Joslyn, and G. L. Marsh. Fruit and vegetable juices. 549 p. Avi Publishing Co., Inc., New York, N. Y. 1939. 84 Amerine, M. A., and W. De Mattei. Color in California wines. III. Methods of removing color from the skins. Food Research 5(5):509-19. 1940. 85 Berg, A. Color extraction for port-wine manufacture. The Wine Review 8(1) : 12-14. 1940. 78 University of California — Experiment Station juice is removed for fortification, while the remaining pomace is watered and used for the production of distilling material for fortifying brandy. If presses are used, the continuous press is commonly utilized because of its convenience and ease of operation, although slightly better musts for fortification are obtained with a basket press. FORTIFICATION 88 Time. — The sugar content will be reduced to the proper amount for fortification in 2 to 6 days, the time required depending on the temper- ature, amount of yeast, and other factors. If the rate of fermentation is rapid, the must should be pressed at a sugar content at least 2° Balling above that at which it is desired to fortify the must. The time required for the alcohol tests and pumping is sufficient to permit the Balling to reach the desired degree. Sugar Content. — In California wine makers usually wish to produce a red dessert wine of 20 per cent alcohol and 6° Balling. This means an actual extract content in the dealcoholized wine of about 10 to 12 per cent. The proper Balling at which to fortify in order to secure this ex- tract content depends on the original Balling of the must. The higher the original sugar content the lower the Balling at which the fermenting must may be fortified in order to secure a finished wine with an extract content of 10 to 12 per cent soluble solids. Some approximate Balling readings for fortification of grapes harvested at different degrees of maturity are given in figure 1 (p. 49). The data in this figure were taken at the time of fortification and may be somewhat lower than winery figures. Fortifying Brandy. — In California, brandy of over 180° proof is commonly used for fortification of red dessert wines. According to da Costa 87 and Garino-Canina, 88 the best dessert wines of Portugal and Sicily are produced when the fortifying brandy is of low proof, specifi- cally 152° to 156°. Such a brandy contributes a greater degree of aroma to the wine both before and during the aging and is said to blend with the wine better and to add to its quality on aging. Wines prepared in this fashion, however, must be aged for a longer period of time in order to attain their best quality. According to de Castella : The spirit used in fortifying Port is a matter of vital importance, and has con- 86 See also p. 46 to 58. 87 Costa, L. Cincinnato da. Le probleme des eaux de vie et des alcools — Influence de leur provenance et leur degre dans le vinage. V en,e Congres International de la Vigne et du Vin. Rapports, Tome 2, Oenologie. p. 181-89. Editorial Imperio, Lis- bon, Portugal. 1938. 88 Garino-Canina, E. Sulla valorizzazione delFalcool de vino sotto il riflesso economico della tecnica vitivinicola. Regia Stazione Enologica Sperimentale di Asti Annuario (Serie II) 1:215-21. 1934. Bul. 651] Commercial Production of Dessert Wines 79 siderable bearing on the character and bouquet of the wine. In the best Douro vineyards a highly rectified silent [neutral] spirit is never used — only a spirit distilled at fairly low strength, and which retains a good deal of the flaror of the wine from which it is made. Its strength when added to the wine is usually 37% over proof [156.4° proof in the United States]. This spirit is often distilled in a pot still, and can, of course, only be made from sound wine. . . . The type of spirit used, and the fact that it is not silent spirit, are points of great interest. ... It is noteworthy that this special spirit is only used for the high-class wines, which take a long time to mature. . . . For cheap wines, which are to be consumed young, or for slightly increasing the strength of an old matured wine, should such prove necessary, silent spirit, highly rectified, is used. The merchants of Oporto are most particular in selecting their fortifying spirit. . . . This spirit is, and has been for many years, exclusively made from wine. 89 Brandy of over 170° proof is said to have a bad influence on the color when fortification takes place on the skins. Before Prohibition there was a widespread prejudice in California against the use of fortifying brandy of proofs of 180° or over due to the reduced color and quality which was supposed to result from their use, and spirits of 170° to 180° were commonly used. We have been unable to confirm this prejudice adequately. Method. — Government regulations require that wines be placed in a fortifying room (see p. 46) for fortification. The wine to be fortified is therefore pumped from the fermenting room to the fortifying room and the total volume determined by the United States gauger. The alcohol content is determined by the gauger, who also determines the proper amount of brandy required to bring the alcohol content to the desired percentage of alcohol (to 19.5 to 21.0) . This amount of fortifying brandy is then weighed out and pumped into the fortifying tank. The amount required is calculated by one of the formulas given previously (p. 52) and by reference to the Gauging Manual (cited in footnote 69, p. 54). A record of these amounts and percentages must be kept on the proper government forms. In order to stop the fermentation immediately, the added fortifying brandy and must should be promptly and thoroughly mixed. Because of the great difference in specific gravity of the alcohol and the wine, mechanical mixing will be required, especially if the fortifying tanks are large. Continuous pumping over is sometimes used to do this. Re- lease of compressed air in the bottom of the fortifying tank has also been used successfully. The alcohol content of the wine is then redetermined and the wine pumped to the storage cellar. Note: All determinations and calculations of the gauger should be 80 Castella, F. de. Port. Victoria Department of Agriculture Journal 6:176-91. 1908. 80 University of California — Experiment Station cheeked before fortification. An easy and common method of preventing large errors is always to fill the fortifying tanks to the same height. Experience will then indicate the approximate amount of fortifying brandy required, and gross errors can be eliminated. AGING Aging of red dessert wines involves the removal of the lees from the wine and subsequent aging of the wine to the required degree of oxida- tion and color. The high alcohol content of dessert wines usually pre- vents undesirable bacterial activity in the lees, but racking the wine from the lees as soon as it falls clear is generally considered prudent. This is particularly advantageous where the wine is contaminated with bacteria, for certain bacteria are able to act in the presence of a high alcohol content (see p. 144) . In general, the wine should be racked within a week after fortification and again in the storage cellar within a month. The lees material is usually diluted and used for the distillation of forti- fying brandy or for lees brandy. Ordinary-quality wines are stored in large-sized redwood or concrete containers. The best wines are stored in smaller-sized oak containers, for example, in Portugal in pipes of about 140 gallons' capacity. Treatment. — Under the best conditions — good-quality grapes, clean fermentation, and safe and sufficient storage — the red dessert wines be- come almost brilliant naturally, and clarification or other treatments need only be considered just prior to shipment if at all. Wines which do not become clear by themselves within the first six months should be clarified. The usual sequence for such red dessert wines involves cooling to about 20 °F either by pumping through a heat exchanger or by the storage of the wine in a room of about this tempera- ture. The wine is left at this temperature for about one week and then is filtered into the storage tanks. It may be fined and filtered before or after such treatment, usually with bentonite or gelatin. If the wine is diseased, various treatments may be necessary (see p. 147). There has been a large demand since Prohibition for cheap red des- sert wines. Early maturation is necessary, since these wines must be delivered to wholesale buyers when young. The treatment of the wines for this purpose may involve heating to 110° F for a period of about a week, or electrolytic treatment for a day or two, as well as the measures outlined above for wines difficult to clarify. Additional oxidation by the use of oxygen may also be used advantageously in some cases. There is an economic advantage in moving ordinary-quality wines rapidly, which justifies such practices. High-quality wines should, of course, not be so treated. (See p. 61.) Bul. 651] Commercial Production of Dessert Wines 81 Storage. — The time required to bring a red dessert wine to its optimum maturity varies with the type of wine, the temperature of storage, the size of container, and the treatments which the wines have received. Red wines of high tannin require a longer period of aging. Wines forti- fied with low-proof fortifying brandy also require a longer period of aging. The rate of aging is, however, considerably speeded up at warm temperatures. DIRECTIONS FOR MAKING SWEET, WHITE DESSERT WINES 90 The chief types of white, sweet, nonrancio-flavored wines produced in this state are Angelica, muscatel, and California white port. Forti- fied sweet wines are occasionally produced in eastern United States from Catawba and other non-Vitis-vinifera varieties by similar pro- cedures or by blending. Angelica. — One of the early procedures used in southern California, according to Hyatt, 91 for the production of Angelica was to fortify 3 gallons of must with 1 gallon of brandy. Assuming a proof in the forti- fying brandy of 140° to 160°, which was the usual range with the pot stills used in those days, the finished wine probably contained from 17.5 to 20.0 per cent alcohol. In the period just before Prohibition, commer- cial practices approximately followed this tradition, and unfermented or nearly unfermented free-run juice was fortified; but brandy of high proof (170° to 185°) was used for the fortification. Post-Repeal prac- tices have permitted a longer period of fermentation before fortification, but Angelica is still one of the sweetest of the ordinary commercial types of wine (see p. 24). It is interesting to note that the Malaga wines of Spain are also very sweet, but their high sugar concentration is ob- tained by the use of boiled-down must and the wine is much darker in color than a California Angelica. Varieties suitable for Angelica are given on page 37. Muscatel. — Muscatel is a fortified wine possessing a distinct muscat flavor and having over 10 per cent sugar. It has been produced in Cali- fornia since the early days but only achieved a large-scale demand during the period just after Prohibition. Its aromatic and distinct varietal flavor is easily recognizable and apparently has strongly im- pressed the new wine drinkers of the early post-Repeal period. It is not generally considered a quality sweet dessert wine because of its sweetly mawkish character. Special muscatels, such as those made from Muscat 00 General references on this subject in addition to those given in specific foot- notes in the section are listed on p. 173. 91 Hyatt, T. H. Hyatt's hand-book of grape culture. 2d ed. 279 p. A. L. Bancroft & Co., San Francisco, Calif. 1876. 82 University of California — Experiment Station Canelli, offer some promise of improving the quality of this type of wine. A list of muscat-flavored varieties is given on page 36. White Port. — California white port is a type which was only occa- sionally produced in the pre-Prohibition period. It is a very light- colored, white, sweet wine, the water-white appearance being due to the method of treatment used during the finishing process. In the pre-Pro- hibition period, the decolorization was usually made with bone charcoal, but now activated vegetable charcoal is commonly used. Varieties suit- able for Angelica are generally satisfactory for this type of wine. Slightly pink wines produced from varieties such as Grenache, Carig- nane, and Mission can also be used, for they lose their color during the finishing process (see p. 37) . Owing to the present legal restrictions, wine to be converted into white port must be designated as such at the time of fortification since only this wine is permitted to be decolorized. Since, however, white-port stock may be marketed as Angelica, it is the custom to carry Angelica stock as white port on the winery records. HARVESTING AND FERMENTING The picking and handling of grapes for white dessert wines is no less important than that of white table wines. The grape varieties used are usually riper and the resistance of the berry to injury is therefore reduced. The grapes should always be picked with the least possible injury. Rotten or moldy grapes are higher in enzyme content 92 than sound fruit and should not be picked, if light-colored, fruity, clean, white dessert wines are to be produced. Bruised fruit also frequently yields wines with an undesirable brown color. Every effort should be made to reduce the period from picking to crushing to the shortest possible time interval. To achieve this, it is important that wineries should arrange picking and transportation schedules with the growers so that the crushing facilities of the winery are not overcrowded at any time by deliveries. One- and two-day delays in unloading trucks (sometimes seen in California) are unnecessary, reduce the quality of the wines, and do not make for good will with the growers or the transportation companies. Undesirable bacterial action and harmful metal contamination are the further penalties for such delays. The metal contamination is liable to be especially serious for white dessert wines if gondola trucks are being used. In general, iron gondola trucks are not advisable for transporting grapes for wine making. "Laborde, J. Cours d'oenologie. Vol. I. 344 p. (See especially p. 231-33.) L. Mulo, Paris, France. 1907. Bul. 651] Commercial Production of Dessert Wines 83 The picking of raisined fruit may yield wines with a dark color and an unpleasant raisin or caramel flavor. The general impression that only raisined Muscat of Alexandria grapes yield highly aromatic muscatel wines is incorrect. Very early harvesting, say at 21° Balling, does result in wines of low varietal flavor. The real reason for the enhanced flavor of late-picked muscats is because both the first- and second-crop grapes are ripe at that time, but usually the varietal flavor is then accompanied by a strong raisin flavor. The best muscatels would result if the first-crop muscats were picked after they reach 26° Balling and before they raisin too badly. In some districts the Balling reading at the time raisining begins may be even higher than this (at Escondido, in southern California, for example). But unfortunately, in the districts where the greatest acreage of muscats is planted, the high summer temperatures result in sunburning and raisining before all the grapes, the second as well as the first crop, are really ripe. The unripe second-crop muscats would not dilute the flavor of the first-crop muscats if the first crop were picked separately. The grower and winery must therefore choose between three possibilities : too early picking and deficiency of flavor ; picking only the ripe first-crop grapes at their optimum maturity (which is more expensive) ; and late picking of all the grapes and having a darker-colored, more raisin-flavored wine. Some pre-Prohibition win- eries chose the second alternative and then used the second-crop muscats for making muscat distilling material. Crushing. — The usual crusher and stemmer is used and the must pumped to the open redwood or concrete f ermenters. The rotary crush- ers should be operated at a high speed with only a moderate load when shriveled or raisined fruit is being crushed, every effort being made to crack the skins. Practically all wineries stem the grapes at the same time, and this is a desirable practice. Grapes intended for Angelica and California white port are usually separated from the skins shortly after crushing, particularly if the white port is being made from grapes which have some pigment in the skins. Some wineries simply drain the free-run juice for fermentation and wine making and then dilute the remaining pomace with water for the production of distilling material. With inexpensive grapes, when grapes are plentiful, or when there is need of large amounts of forti- fying brandy, this practice is justified; otherwise, the grapes should be gently pressed and the press wine added to the free-run juice, while the press cake only is diluted and used for making distilling material. Muscats are left in the skins from 6 to 36 hours after crushing before pressing, in order to extract more of the flavor from the skins; pressing is also easier after a period of fermentation. The time on the skins 84 University of California — Experiment Station should not be too long else too much tannin and coloring matter will also be extracted. Conduct of the Fermentation. — Sulfur dioxide is added to the juice or to the must as soon as possible after crushing. If the pressing is to be delayed more than a couple of hours, the sulfur dioxide should be added before pressing. The amounts to add depend on the condition of the grapes and the temperature of the fermentation (see p. 42). At least 75 parts per million are ordinarily added. Potassium metabisulfite, sul- fur dioxide, or a 6 per cent solution of one of these may be used. The sulfured must should not be allowed to come in contact with any metal surfaces. (See also Bulletin 639, p. 39-45, 88.) Pure yeast cultures of one of the common strains are sometimes added about an hour after the sulfur dioxide. From 1 to 3 per cent of the pure culture is commonly used. For producing white, sweet dessert wine, with its restricted period of fermentation, there is little advantage to be obtained from using a particular strain of pure yeast. (See p. 43 for the possible effects of mixed cultures.) After pressing, the fermentation may be conducted in closed or open containers. Closed tanks are preferable because they reduce the possibil- ities of contamination and overoxidation. Special control of the temper- ature is not ordinarily required unless the initial temperature of the must is high or the fermenters are of too great a capacity and have a low rate of heat loss. The products of fermentation at a moderate tem- perature (below 80° F) are considered more desirable than those of a hot fermentation, and the resulting wine is less apt to be contaminated. When the wine reaches the appropriate Balling, it is pumped to the fortifying room. SPECIAL PROCEDURES FOR MUSCATEL Improvement in California muscatels can undoubtedly be brought about by the use of better varieties (see p. 36). But the extraction of the maximum muscat flavor of which the common Muscat of Alexandria variety is capable is occasionally desirable. The simplest method of in- creasing the muscat flavor in the wine is that of prolonging the period of fermentation on the skins. This period cannot be unduly prolonged because the must darkens when fermented on the skins, too much tannin is extracted, and there is greater possibility of bacterial spoilage in the cap. Punching the cap down or pumping the free-run juice over will also increase the extraction of flavor. The possibility of overoxidation is likewise increased, but if the punching-down is not continued too long, it may not be serious. Leaching of the skins with the free-run juice is Bul. 651] Commercial Production of Dessert Wines 85 a similar type of procedure. In some cases, the skins are put through a special macerating machine before the leaching. These machines (see fig. 12) roll the skin over a metal surface so that the flesh is completely pressed away from the skin. If the additional flavor is worth a considerable amount of labor, heat treatment may be used. The free-run juice is removed from the skins and heated to above 140° F. The heated juice is then poured onto the skins and allowed to stand for several hours before pressing. The muscat flavor is increased, the pressing is easier, and the heat treatment acts as a partial pasteurization and insures a clean fermentation. The latter consideration is important, since muscat musts are traditionally diffi- cult to ferment and quickly spoil, even during fermentation. There is some evidence, however, that the pectin and gum content is unduly in- creased by this procedure and the wine may be somewhat difficult to clarify. All such treatments, involving either heat or maceration of the skins, should be used with caution until comparative tests establish their influence on the flavor, the collodial condition of the resulting wines, and the difficulty of clarification. FORTIFICATION In the best fermentation procedures for making white sweet dessert wines, the skins are not in contact with the must for a very long period, even with muscatels. The must is therefore free of skins during most of its fermentation and has simply to be pumped to the fortifying room at the proper time. Angelica material may be moved to the fortifying room soon after pressing, the period of fermentation depending on the degree of sweet- ness desired by the wine maker and on the sugar content of the grapes. Muscat and white-port musts are fermented longer before fortification. The approximate Balling to fortify musts in order to secure the proper Balling in the finished wine is given by the chart in figure 1 (p. 49). Allowance must be made for changes due to fermentation during the pumping into the fortifying room as well as during fortification and mixing. The Balling reading may drop 2° or 3° during this period, if the must is in full fermentation at a favorable temperature. Experience with plant conditions and equipment and with different gaugers' methods and speed will establish the proper degree of safety which is necessary to secure the required sugar in the finished wine. As previously mentioned for port, it is advantageous to fill the forti- fication tanks always to the same level. The gauger's and wine maker's calculations of the amount of fortifying brandy necessary can then be more easily checked with the experience of previous fortifications. The 80 University of California — Experiment Station Fig. 12. — A, Outside view of Garolla skin macerator showing brushes which are used to keep the holes open ; B, inside view showing the paddles to keep the pomace broken up as the outside cylinder revolves. BUT, G51] COMMERCIAL PRODUCTION OF DESSERT WlNES 87 wine maker or his chemist should always determine the alcohol content of the fermenting must simultaneously with the gauger and should also calculate independently the amount of brandy needed. This check will prevent large and serious errors. The required amount of fortifying brandy is then pumped into the fortifying tanks and mixed well with the must. Compressed air or pumping over are common methods of mixing. The use of neutral fortifying brandy is recommended for Angelica and white port, but brandy made from muscat distilling material should be used in the fortification of muscatels. Fortifying brandy of 170° to 180° proof made from a fresh, dry, muscat wine has a fruity, muscat aroma and will enhance the aroma and varietal flavor of the musts which are fortified with it. AGING Neither Angelica nor muscatel nor California white port are aged for a long period in California. The main problem in their aging is to allow them to lose their alcoholicity and become brilliantly clear, at the same time preventing them from becoming overoxidized or developing a rancio flavor. Sufficient aging should be allowed, however, with the Angelica and muscatel to permit them to become smooth. This usually requires three or four years. The better muscatels should receive an even longer aging. The best muscatels at Frontignan in France, which are similar to California mus- catels in sugar content although somewhat lower in alcohol content (17 to 19 per cent), are frequently aged from five to ten or more years in puncheons. The time of aging will depend partially on the size of container, the type, composition, and quality of the wine, and the tem- perature of storage. In general, the larger the container, the better the wine, and the lower the temperature of storage, the longer the period to bring the wine to maturity. Wines of low quality, those in small con- tainers, and those stored at high temperatures, mature more rapidly. Racking. — The wine must usually be pumped from the fortifying room immediately. Within a week it should be racked off of the heavy yeast sediment and moved to a cool part of the cellar. Sound, clean, properly fortified dessert wines made from good healtlry grapes with a clean fermentation will clarify themselves rapidly. By the first of the year or earlier, the wine will be clear and should be racked again. Peri- odic racking every six months is recommended for the first few years. Filling. — Dessert wines should not be allowed to remain in unfilled containers for long periods of time. While the necessity of regular filling is not as urgent with dessert wines as with table wines, the containers should be regularly filled, particularly with the white sweet dessert 88 University of California — Experiment Station wines, where the development of large air spaces will lead to undesirable darkening of the wine. Filling every month the first six months and once between rackings thereafter is considered good practice, if barrels or puncheons are used. Early Finishing. — With large inventories of ordinary dessert wines at the end of the vintage season, the wine maker faces the problem of disposing of those wines which will not benefit by the usual aging process. This involves the finishing of the wines for bulk shipment within the year after they are made. The commercial demand for young, cheap, ordinary-quality wines has been a problem for wineries producing dessert wines as well as for those producing table wines. Early stabiliza- tion and shipment of these wines is necessary. The whole procedure, which is recommended only for such wines, usually involves reducing the temperature to slightly below 20° F for about a week, filtering, pumping through a heat exchanger to 160° or higher, and possibly fin- ing with bentonite and close-filtering. It is sometimes necessary to re- peat the treatment to obtain permanent brightness. This type of wine is also frequently kept at a level of 50 to 100 parts per million of sulfur dioxide to prevent growth of certain organisms. (See p. 144.) The best muscatels and Angelicas, however, are simply racked, and after a year or two stored in small containers of 54- to 200-gallon ca- pacity. They are bottled after two to ten or more years' aging. The wine should be fined during a cool period. Wines which do not clear up by themselves should not be aged in this fashion and should be handled by the rapid-finishing methods given above. Preparation of White Port. — The present-day California white port is made by adding from 1 to 10 pounds of one of the commercial activated carbons to 1,000 gallons of white or nearly white wine. Both "vegetable- char" and "bone-char" decolorizing carbons are used; the latter are preferred by most wine makers. The minimum quantity required to decolorize each lot of wine should be determined in the laboratory, because it varies for each wine and for each type of charcoal. To obtain a light straw-yellow wine, about 2 to 4 pounds of the newer types of activated bone charcoal per 1,000 gallons will suffice; for water- white port, from 5 to 10 pounds or over per 1,000 gallons will be required. The wine is filtered off the charcoal after several days and is usually shipped or bottled after a short period of aging, but before picking up any color. Practically all the grape flavors and aromas are removed by this treat- ment, and the wine becomes practically water-white in appearance. If too much charcoal is used, the wine will take on a foreign undesirable taste : the activated carbons induce the oxidation of alcohol and result Bul. 651] Commercial Production of Dessert Wines 89 in a rapid increase in acetaldehyde content, which may reach 300 parts per million. Furthermore, excessive use of charcoal results in formation of white deposits (see p. 149). But since the charcoal also removes off- flavors, wines having a bad odor or taste, which would otherwise be unusable, sometimes can be treated with an activated charcoal and sold as white port. Directions for determining the minimum quantity of carbon neces- sary are given in booklets distributed by manufacturers. 93 Saywell 94 sug- gests that the wine be heated to 100° to 140° F. Agitation of wine with the carbon is also desirable to increase the rate of decolorization. After decolorization is completed, the carbon in suspension may be removed more easily if the wine is fined with ben- tonite at the rate of 2 to 6 pounds per 1,000 gallons. The bentonite-treated wine is then filtered through a filter press using diatomaceous earth as a filter aid. Gentilini 95 has found very marked differences in the decolorizing properties of Italian charcoals, and Schatzlein and Sailer" 6 have found similar differences in German charcoals. In these tests, some charcoals were found to be very inferior in decolorizing properties, while others added large amounts of calcium or other ash constituents to the wine or had inferior absorption properties for pectins. No comparable study is available for commercially available American charcoals, so the wineries in this state should conduct their own tests to determine the most efficient products. In addition, Motusz" 7 has shown that the ash content of certain ac- tivated carbons is excessively high and that metallic impurities such as iron and manganese enter the wine during their use. He recommends that the content of iron and manganese salts should not exceed 0.01 to 0.05 per cent in the activated carbons used in the winery. The possibility of pickup of excessive amounts of anions, such as phosphates, should also be borne in mind. 93 Anonymous. The modern purifier. 98 p. Industrial Chemical Sales Division, West Virginia Pulp and Paper Company, New York, N. Y. 1937. Anonymous. Conditioning wines with Darco. 10 p. Darco Corporation, New York, N. Y. 1936. Anonymous. Handbook for counter-current treatment with activated carbon. 29 p. Darco Corporation, New York, N. Y. 1939. 94 Saywell, L. G. Use of activated carbons in wines. The Wine Review 3(9): 17-18. 1935. 95 Gentilini, L. Di alcuni carboni in enologia. Conegliano Regia Stazione Speri- mentale di Viticoltura e di Enologia Annuario 7:301-15, 317-37. 1937. 96 Schatzlein, Chr., and E. Sailer. Die Aktivkohlen und ihre Verwendung zur Behandlung von Weinen und Siissmosten. Wein und Rebe 18 : 258-65. 1937. 97 Motusz, J. Extractable mineral components of active carbon preparations. [Translated title.] Kiserletiigyi Kozlemenyek 38:161-72. 1935. Abstracted in: In- dustrial and Engineering Chemistry, news edition 14:87. 1936; and in: Chemical Abstracts 30:1954. 1936. 90 University of California — Experiment Station DIRECTIONS FOR MAKING SHERRY AND OTHER RANCIO- FLAVORED WINES 98 Flavor. — Sweet white wines of low acid and tannin content, contain- ing appreciable quantities of unfermented sugar, readily develop, on exposure to air, a peculiar characteristic taste known as "rancio" (gout de ranee in France). This taste is produced largely as the result of the oxidative caramelization of the sugars and oxidation of other substances present. This oxidation and caramelization is not desirable in delicately flavored, fruity wines like muscatel, but when suitably controlled it is desirable in the sweet Madeira, Malaga, and Marsala types of wine. (See p. 31 and p. 108.) The sugars present in the wine may also slowly caramelize during storage, particularly at higher temperatures. The flavor developed in California sherries of commerce by suitable treat- ment is often, but improperly, called "sherry taste," although it is not at all analagous to the flavor of Spanish sherries. It does resemble the flavor of certain wines produced in the south of France — Banyuls, for example — and in Madeira. The excessive caramelized flavor of certain California sherries is not, however, a true rancio flavor. The production of the proper rancio flavor requires considerable skill and care. Another feature these wines have in common is that their flavor is greatly improved by storage in oak, even at high temperatures, for they are tolerant of oak extractives. The rancio-flavored wines im- prove more during storage in cask, even when this is prolonged, than do other wines, and (with the possible exception of the more delicate ftno types to be discussed later) they do not become "vapid" on exposure to air. Acetaldehyde Production. — Oxidation of alcohol to acetaldehyde is an important factor in the production of flavor, both in Spanish sherry and rancio wines such as California sherry. Acetaldehyde is a particu- larly important flavoring constituent of the fino type of Spanish sher- ries, where it is present in relatively high concentrations. Trillat 9 " reported the acetaldehyde content of a fifteen-year-old Jerez (Spanish sherry) wine as 276 mg per liter of free aldehyde and 418 mg per liter of total aldehyde. Rocques 100 found a Jerez wine to contain 199 mg and an amontillado type 383 mg per liter of total aldehyde. Normal table wines (both red and white) usually contain less than 50 mg per liter even when old, and the acetaldehyde flavor in them is decidedly dis- 98 General references on this subject in addition to those given in specific footnotes in the section are listed on p. 173-74. !,! ' Trillat, M. A. L'aldehyde acetique dans le vin, son origine et ses effets. Institut Pasteur [Paris] Annales 22:704-19, 753-62, 876-95. 1908. aoo Eocques, X. Le bouquet des vins. Revue de Viticulture 12:95-99. 1899. Bul. 651] Commercial Production of Dessert Wines 91 agreeable when it reaches 100 mg per liter. Rancio-flavored dessert wines may and do contain more acetaldehyde than the table wines without becoming objectionable to taste, particularly when they also contain sufficient sulfite or other substances to combine with it. Combined acetal- dehyde blends into the wine better than free. The characteristic flavor of Marsala, Madeira, and sherry wines from Europe is obtained largely by a process of manufacture peculiar to each wine : the use of caramelized grape concentrate as a base for Marsala wine, the use of a cooked wine as a base for Madeira, and the use in Spanish sherry of a peculiar yeast capable of vigorously fermenting must in its anaerobic stage, and of forming a heavy film in its aerobic or oxidative stage. In California, on the other hand, no uniform process is used at present in making any of these types. Because of this differ- ence, the European as well as the California practice is given in this section. MAKING SPANISH SHERRY WINE Types. — A number of distinctly different types of sherries are pro- duced in and about Jerez de la Frontera in Andalusia, a province in the south of Spain. These wines are of pale or deep golden color, sweet or dry, and with an alcoholic content of 14 to 24 per cent, usually 16 to 18 per cent, and from 2 to 20 per cent of extract, averaging about 5. These Jerez (Xerez), or sherry, wines are each of surprisingly standard and uniform market grade. 101 This standardization is obtained by a process of blending different wines and wines of different years, according to qualities acquired or not acquired with time, in a solera system of ma- turing. Certain of them are also subjected to heating in the sun in oak containers for varying periods of time, although this is not the case for the drier, finer, and more characteristic wines. Several of the sweet wines of Jerez, used primarily for blending, have a characteristic rancio flavor, and this has resulted in the confusion of rancio flavor with the true sherry flavor. The basic types of sherry wine in commerce at the present time in the sherry district proper, according to Gonzalez Gordon (cited in footnote 16, p. 15) are the finos, amontillados, and olorosos. The finos are very light straw-colored wines having a peculiar, slightly bitter, hazelnut flavor. They are practically dry (as low in reducing sugar content as the dry table wines of commerce). They are free from a .rancio flavor, but may have a slight and pleasant oaky aroma and taste. 101 Greenup, Julian C. The Spanish spirits and wine industry. Special report by the Acting Commercial Attache, Madrid, Spain. October 7, 1933. 20 p. (Type- written manuscript distributed by Wine Institute.) Daniel M. Braddock. The wines of Spain. Voluntary report of the American Vice-Consul, Barcelona, Spain. May 16, 1933. 58 p. (Typewritten manuscript dis- tributed by Wine Institute.) 92 University of California — Experiment Station Their peculiar flavor is developed largely as a result of their secondary fermentation. In the nomenclature of the cellar foreman at Jerez, they are sometimes called palmas. The amontillados are somewhat darker than the finos, being more yellow or even slightly amber. The name is derived from their resem- blance to the montilla wines which are produced in the province of Cordova. Their character is a result of the long aging of a wine of the finos type. Their alcohol and extract contents are somewhat higher than those of the latter. Vino de Pasto is a wine of the amontillado type but is less dry and of lower quality. The olorosos are wines of an even deeper color, a full amber. With a few exceptions, they are fairly sweet, 1 to 4 per cent sugar, and they range from 18 to over 20 per cent alcohol. The palo cortado is younger wine of the oloroso type and is said to have an aroma resembling that of an amontillado. Amoroso is another oloroso type having a mild char- acter. Olorosos are sometimes called "golden" or "East India" sherries in commerce. The rayas are really wines of the oloroso type but have a greater body, a darker color, and are not as clean (limpio) as a true oloroso. Their flavor is rich but somewhat coarse and they are mainly used for blending. A distinctive wine of the sherry type called manzanilla is produced near Sanlucar de Barrameda. It is dry, and has a pale straw color similar to that of a fimo. The manzanilla aroma is delicate and characteristic. Although of fairly low alcohol content when young, manzanillos range up to 21 per cent or even more when aged. Montilla is a wine resembling sherry produced in the Los Moriles dis- trict in the province of Cordova, According to Gonzalez Gordon, they resemble but lack the finesse of fine amontillados. All shippers of sherry do not follow the same usage with respect to these names. Some of these terms, for example, are used by the shippers as proprietary names in this country, and the wine itself may be a differ- ent type from that indicated by the label. The analytical data for the different types given in table 27 are supposed to be from reliable sources in Spain. Although the data do indicate certain differences, the real classification of the wines into types at Jerez is by means of tasting. Fermentation™ 2 — The best sherry wine is made from Palomino grapes grown on albariza soil (a very white soil containing from about 35 to 40 per cent lime). The grapes are harvested at 12.5° to 14.0° Baume (22.6° to 25.3° Balling) and are then partly dried in the sun 24 to 48 hours on 102 Although a number of popular accounts of the methods used in making sherry wines in Spain are available, such as that of Allen (Allen, H. W. Romance of wine. 264 p.; see p. 134-63. E. P. Dutton and Co., New York, N. Y. 1932.), there are very few authoritative descriptions, and no complete chemical and micro- Bul. 651] Commercial Production of Dessert Wines 93 round esparto grass mats. They are lightly crushed into lagars — stone vats about 2 X 12 X 15 feet — particular care being taken to avoid ex- traction of tannin. Burnt yeso, a natural earth composed of about 90 per cent of calcium sulfate, plaster of Paris, is added to the grapes in the lagar and additional amounts to the grapes in the press (at the rate of about 2 pounds per ton of grapes or about 1% pounds per 100 gallons) . The calcium sulfate precipitates the tartrates as the insoluble calcium tartrate and replaces them by an equivalent amount of the rather bit- ter potassium sulfate according to the following equations, the reaction in a given case depending on the ratio of calcium sulfate to tartrates present: 103 CaS0 4 + 2KHC 4 H 4 6 =CaC 4 H 4 6 + H 2 C 4 H 4 6 + K 2 S0 4 . 2CaS0 4 + 2KHC 4 H 4 6 == CaC 4 H 4 + K 2 S0 4 + H 2 S0 4 . According to Garcia de Angulo (quoted by Gonzalez Gordon, p. 117, cited in footnote 16, p. 15) the yesoing increases the titratable acidity of the must and reduces the pH. This is said to result in cleaner fermenta- tions and possibly also influences the organoleptic character of wine. In addition, the potassium bitartrate content is reduced so that there is less danger of tartrate precipitation later. The mineral content is in- creased, however. The first pressing and free-run juice is run into oak butts of 108 Imperial gallons' capacity (130 United States gallons), in which it is fermented. According to early writers, the fermentation was allowed to proceed naturally, no sulfur dioxide or pure yeast being used to control the fermentation. More recently, however, the use of sulfur dioxide has become common, and de Bobadilla 104 has shown that in Jerez the use of sulfur dioxide is desirable in musts to be used for sherry. The initial fermentation of the other types of sherries is conducted in a similar manner. The grapes used for the dry manzanilla wines are picked at a slightly lower sugar content and are heaped in deeper layers during sun-drying so that less drying occurs. Special Types. — Pedro Ximenez (or Pedro Jimenez) is the exceed- biological study of the process has been published. The account here is based chiefly on the following treatises: Castella, F. de. Sherry: its making and rearing. Victoria Department of Agri- culture Journal 7:442-46, 515-28, 577-83, 621-30, 724-27. 1909. Rocques, X. Les vins de liqueur. Le vin de Jerez. Revue de Viticulture 19: 501-5, 570-73, 594-98. 1903. Gonzalez Gordon, M. M. a (cited in footnote 16, p. 15). 103 Actually only the calcium ions and tartrate ions participate in the reaction to yield the insoluble calcium tartrate ; and potassium, hydrogen, sulfate, and bitartrate ions, together with residual calcium and tartrate ions, remain. 104 Bobadilla, G. S. de. L'emploi de Panhydride sulf ureux au point de vue des qualites organoleptiques et hygieniques des vins. Le point de vue Espagnol. V eme Congres International de la Vigne et du Vin. Rapports, Tome II. Oenologie, p. 80-85. Lisbon, Portugal. 1938. 94 University of California — Experiment Station s M W W m K CO S PL, m CM CC cq M < > HS oj t~- *o t-- ■* lO iO l> "O te 10 o „ rt ^ _ o I I I I I I go ^h © 10 10 co ,-( CM CO CO 00 O O O O O CO r-1 I I I I I I N M (N tO "O ^NMNMNH oo s o o m 1-1 i-H r-l CM .^ < < < m o 6 -3 .a "3 13 T3 .2 -9 ° * 3 S « ■? ° ^ H O A IJallS* 1 1 1 nil 0) Q) Qj vfli * r« ^ (m &> C «B O cc a o Ph < U 2 S? o* I si I J a 4) 03 -2 s ■2- £ S. 5- 5 .6 S O S 2 . e « n s s S s "« *; 5 O 8 5 K) °5 . e Qs^^OOh^O o^ooo •5 O rO •2 l| £ c>0 Gq 6q ^ II I '8 g « 3 2 s fc: o © •S s R 03 S 5 8 £ § •§= o S. c rg a N oq < < < < < m m -a .a a a> r a t, +i : 9 "3 : : : : . J M . . ■ • 3-i 2 : M B : m "• a ^j -a - : o> : c D - 4J ■ T3 . O 4) ra ^S e< <-! . +a m a m 2 2 .S S3 Bul. 651] Commercial Production of Dessert Wines 121 u 831 3 1*8 © o o t; © o g g O 03 o o J2 ,5 pd « £ Pd Pn Ph 9 S O) o HH- Pd « Ph ^ O « PL, Ph Ph O 3 0) Q c ,3 mo, „ 03 © .2 oil •c 2.S « a , © 03 3 £ a, a 3 O ©•« O «i § 2 O S U Ph J O a g £ A^Ctf a -a g CO CS CO o :> -3 a, 33 d ~ 5 ^ 3 S H > > "5 5* ? S £ s 8 § IS | 3 3 -« K ~ © s~ o S OhN .~2 &H ! £4j a s c 3 25 "fi a a c o 3 -<3 O Ph 0"Ph s s o o Pd Pi 3 I 2 3 a ° s a s d >>£ C O >- T3 © s- o © co 5c > © O 03 03 a Ph CO CO CQ '§ S S § 3 « >> ^ -2 & ^ ^ 02 02 H > b a ii ■2 :% a-, ^ •§ tt, *-* s- O *» e ~. ^3 "e 3 2 ^3 ^ « 3 o ^5 O CQ K* N ofcS -s a is i . la>>3 3 fi-SS > ap a flc3^ s -3.2 a. S to. SO fe HJT3 . "^ 33^b2 05 "43 *h a-^T3 o to' - "- "-o 03^^03^ ? fl,o w "£ ■§2 I 03 . >, <; co 03 O So O 3 T3 H • 2^ S -S 3 b 2 >i.3 =3 be o.a-ogoj o s o e"2"1 S .co a, ¥X w2-a c^" - *-■ § 5 o^Ph So§5li ^ 2 o3.2 N 2 - ^-s » — b ^ >- © 3-3 ° o3 © ** o3 co to 3 a § f s- ©•^ ^ $ ■^.aj 2"3 W » , -S S3 H »^J S3 ® © "43 S 3 ■£ g g ^02 o3 ^•2-S dfa Sills co co_- 43 g o co^ 43 n Ph^ % * « 3 122 University of California — Experiment Station flavor predominates over that of others. There must be a harmonious blending of the bitter with the fragrant and sweet herbs. Prolonged extraction of the herbs is undesirable because the more slowly dissolv- ing astringent and bitter flavors from the herbs will give the wine peculiar objectionable flavors. The vermouth mixture should be tasted periodically during extraction, and at the first sign of any peculiarity the wine drawn off the herbs and filtered. This process is used in making vermouth in Tuscany, but in Turin an alcoholic extract of herbs and drugs is preferred. Here the mixture of herbs is immersed in 85 per cent alcohol, 10 liters to 6 kg of herbs, and allowed to infuse for 8 days. The extract is then drawn off the herbs and mixed with 18 liters of fresh alcohol and 7 liters of common white wine, and the 35 liters of mixture concentrated by distillation to 18 liters. This residue in the distilling flask is cooled and allowed to stand 15 days, being stirred every day. It is used in flavoring the vermouth base at the rate of 1% to 2 liters per hectoliter of base. Quantities of Herbs to Use. — The quantity of herbs and drugs used in making Italian vermouth varies usually from 0.7 to 1.1 kg of mixed herbs or their equivalent of extract per hectoliter of wine base, although quantities as low as 0.4 kg and as high as 2.4 kg have been suggested (1.0 kg per hectoliter corresponds to 1.27 ounces per gallon) . The number of herbs in the mixture, the kinds, and the relative proportion vary greatly. Typical formulas for the herb mixtures, in grams of each kind to be used per hectoliter of wine (26.4 gallons), are given in table 30. None of the published formulas yields a vermouth with the flavor of the im- ported Italian vermouth, although some of them produce palatable and flavorful products. Several prepared mixtures of plant extracts and essential oils are available for use by the neophyte. The more experienced may start with one of the formulas given in table 30 and modify it to suit the taste. The better vermouth producers abroad are very particular about the source of their herbs and the particular mixture used. For example, cinnamon (Laurus cinnamonus) is available from several countries, including China and Ceylon, but that of Ceylon is considered to have the finest flavor and to be the most desirable for vermouth. In a similar way qualities of chincona are available. For the best results not only the best-quality herbs but only those in good condition (age) should be used. Because of the current European war, stocks of some herbs which are secured from Europe are certain to be low, and the prices may rise unduly. American companies should investigate the possibility of grow- ing certain of the most necessary herbs, such as coriander and worm- wood. Furthermore, the extensive flora of aromatic plants endemic to Bul. 651] Commercial Production of Dessert Wines 123 giO>COO ooo «oo gNNXJlO COW • t- UO :g I :m :l :g :g : :§ :g goowo oo • •© g -SS, : :~ :g : :28 : : g -OOOO O iCOOOOOO g .©10(00 O CMOOCOOOO S: © -OOOOOO OOOOO O O CM 5 ■ lO .OOOO^CJMOl ■ io -* © © co o o • oo 5^ § :§2 -H ■ ,-! t-I • >0© 00>0 (MO OCJ •© (M < •oo oo © >o wo g©©«ooo© 000i0«50 • 5 CO U5 m us m o ■cooiNraxj ■< ■ icio ■ *o • © •low .»oo • *o -oo ; ^ ■ oo ■ ■<*< • cm ■ ^ oo • *-i • oo ^h [ t-4 ■ • cm t-h -,-t ■ ■ ■ • CM (MCOCMCM • -eq cm co cm th £: 3 o<§ <5«WOOOOOOWHWOOO^^gZO^a} is 2 g e * 55-2 1 1 S3 >» -8 .g gd+L; -§ ~ a ^a^S S >fi (Vs C . i(3^^ „ CD i g& g Is s ^il a t-i »« W> ;~J3 • S t> - o . I «-£ o 8^ fc'S-csg o w, cog „ g g ^SS-3^" Ss E gs w o e ^^ cd ..3 **> £ 2 3 g a 2 - cd'S^ 3 r^g O Wig § OMj O «GQ ©,9 II § It- a?- - a ^l sSo^s S^ § ^1 Is 8 »2 i a Id^ Sg =5dg-S og .-8 « '".ScnS ^ I ^ S« ^ "£* ^°M dgcScD » g*o ^SSg B-a ■Bl" ^3»33 -§•5 »^^^-|3s |S s a . Sg"^ §8 ?.g2|ai8r. l|s| ill--* §1 Illilll ".Illi lHJll a cd .£ ^=3g ® c- 3 . ^ 3 ^ § sccdcvcd Scccc^awwcoaccco^ccassaa i-I <^-^^" gM«j- MM» MMi.l« » 00 C3 -g CD.S ooo teaS^cDcS OlO-HCONrtMrHrtr-INCONINNCNN ,§ g OflSJ 13 t-Tcm'"© t-"oo"oTo £~£>~co ^"jo co r£oo oj ^ ® O ^ g jjjjj ■Soooodddddoddo odd S ^^^^^^^^^^^^^^^^^ g <1 M 124 University of California — Experiment Station California, the United States, and the Western Hemisphere has so far been but little examined for plants suitable for substituting for Old World plants or useful as ingredients for flavoring vermouth or new types of wine. Sweet Vermouth in California. — Angelica or muscatel wines may be used as a base. A very sweet Angelica, practically a fortified must, is very useful. This should be adjusted with added grape concentrate to the proper sugar content and with citric acid for the proper acidity. Allowance should be made for dilution when extracts of herbs are to be used. The formulas given in table 30 show that vanilla should be used with discretion, if at all, in sweet vermouths — contrary to certain American practices. The newly made vermouth should be fined, aged in wood for a short period of time, and then bottled. Prolonged aging is not desirable be- cause of possible loss of aroma by volatilization and oxidation. Pasteur- ization, refrigeration, and filtration are usually sufficient to stabilize the product, but Luckow 139 reports persistent turbidity, especially at low temperature, if certain types of wood are used, notably licorice (Glycyrrhiza glabra) and catechu. He recommends that if these are used they be extracted separately, diluted, cooled, allowed to settle, and then filtered before they are added to the wine base. Other herbs may yield substances which will cloud the wine. Refrig- eration to a low temperature (15° F) is recommended to start the pre- cipitation of these substances. In Europe, fining with Spanish earths is recommended, hence bentonite should give good clarification here. DRY (FRENCH) VERMOUTH Composition. — The typical French dry vermouth is the highly prized product of Noilly Prat and Company produced at Marseille. This is dryer and somewhat more bitter and pungent than the Italian vermouth. A typical analysis is as follows : alcohol, 18 per cent by volume; reducing sugar, 4 per cent; total acidity, as tartaric, 0.65 grams per 100 cc; volatile acidity as acetic, 0.053 grams per 100 cc. It is easier to prepare a vermouth similar (even if not identical) in flavor to the Italian type than to duplicate the popular Noilly Prat. Two difficulties are experienced : the choice of wine base and the selec- tion and preparation of the proper mixture of herbs. Wine base. — Most California wine makers prefer to use a neutral sauterne-type wine as a base. The sauterne base is prepared by fortifying one lot of California sauterne low in sulfur dioxide content to 24 per cent alcohol and then mixing this with an equal volume of the 12 to 14 per cent alcohol sauterne to obtain a base containing 18 to 18^2 per cent 139 Luckow, C. Trubung in Wermutbitter. Wein und Rebe 19:11-13. 1937. Bul. 651] Commercial Production of Dessert Wines 125 alcohol. This base is sweetened by blending and adjusted to the desired acidity with citric acid. Better dry vermouth would be produced if a wine of higher natural acidity were used. Preparation. — The kinds and amounts of herbs used in published French vermouth formulas are given in table 30, page 123. Under Cali- fornia conditions about % ounce herb mixture, or its equivalent of extract, per gallon of wine base is best. The lack of coriander, cinnamon, and clove, and the high amounts of wormwood in the published formulas for dry vermouth may be noted. A water extract of the herbs is preferable for the French vermouth, although maceration in wine is widely practiced. The flavored wine base is treated as the Italian vermouth but may be aged somewhat longer in wood. Progressive fractional blending during maturation, that is, a solera, should prove useful in developing a French-type vermouth. OTHER HERB-FLAVORED WINES In France a very large number of herb-flavored white and red wines, other than vermouth, are sold. The most popular of these are Byrrh and Dubonnet. They are made with wines from the south of France wines which, in general, have low alcohol and extract contents. These wines are blended with wines containing a small amount of sugar and are fortified to about 18 per cent. The herb mixtures used vary, but chincona bark, which contains the alkaloid quinine, is a major constitu- ent (25 grams or more per gallon). Small amounts of vanilla are also used. The flavors of cinnamon and angostura may be found in others. These wines are very popular as aperitif wines, being consumed without dilution in Europe. Formulas for wines of a similar type for strictly medicinal purposes — such as gentian and rhubarb wines — may be found in the standard dispensatory cyclopedias. CLARIFICATION AND STABILIZATION 140 Wines made from the proper varieties of grapes when handled by the most approved procedures and sufficiently aged almost always reach a brilliantly clear condition without additional treatment. Clarification before bottling is necessary, however, with most wines, and this is par- ticularly true for wines which have been aged in large containers in warm cellars; for wines which are sold without having undergone a complete aging period ; and for wines which are subject to, have, or have had diseases, either bacterial or nonbacterial. 140 General references on this subject in addition to those given in specific footnotes in the section are listed on p. 175—7(5. 126 University of California — Experiment Station The purposes of clarification are to produce a wine which is sufficiently clear for consumer acceptance and which, though it may not remain entirely free of sediment, will stay clear. This latter requirement is particularly difficult when the wine must withstand extremes of heat, cold, and light over a considerable period of time. The wine maker may also wish to clarify the wine in such a fashion as to prevent fermenta- tion or to suppress or eliminate certain organisms which might be the cause of disease. Sterilization nitration, however, is rarely used with dessert wines. While accomplishing some or all of these objects, the aroma and flavor of the wine must be harmed as little as possible. Most treatments arc therefore undesirable for the more delicate wines. STABILIZATION The existing demand for a brilliantly clear wine which will maintain its clearness when exposed to extremes of temperature, to light, and to air, is well recognized by the California producer. It has been necessary to produce a well-stabilized wine that will withstand exposure to the temperatures of the household refrigerators and of the hot air in the retail stores, and also exposure to the light and heat of the sun when placed in window displays. Many of the problems inherent in stabiliza- tion to such influences will be overcome as retailers and consumers begin to learn the proper methods of handling and storing wine, and when more time is given in the winery to the aging of wine. But the present demand for a clear stable product will probably never com- pletely change to one for the more delicately flavored, less-stable old wines, which often contain considerable sediment, especially if not properly handled after leaving the winery. There is undoubtedly a place for both types in the industry. Haziness or cloudiness and the presence of sediment indicate to many consumers a defective product. Such a wine may be fundamentally sound and of excellent quality; although cloudiness and the presence of a sediment may indicate spoilage. Clouding 141 may be due to the growth and activity of bacteria and yeasts; to the formation of insoluble metallic salts of iron, copper, calcium, or tin present in colloidal sus- pension, or as amorphous or crystalline deposits; to the formation and precipitation of insoluble oxidized tannins, coloring matter, or similar products; to the coagulation by heat of certain heat-coagulable sub- stances either originally present or introduced in the process of clarifi- cation; to substances such as cream of tartar or calcium tartrate precipi- tated by exposure to low temperatures; and to other factors. 141 Hall, Lloyd A. Sediment in port wine. Fruit Products Journal 13(9) :270-71. 1934. Bul. 651] Commercial Production of Dessert Wines 127 Present knowledge concerning the nature of the changes involved in the formation of unattractive cloudiness and precipitation of insoluble matters is still rather meager. Furthermore the changes involved in the aging or mellowing of wine are still inadequately understood. Con- sequently no completely rational procedure for the control or elimina- tion of the undesirable changes has been evolved. For the stabilization of common sweet wines, Brown 142 recommends that the newly made wine be racked from the lees, brought to 180° F in a continuous pasteurizer, cooled to 120° to 140° and pumped hot into TABLE 31 Rate of Change in Cream-of-Tartar Content at 24° F (-3.90° C) in Four Wine Types Cream-of-tartar content Storage period In dry white wine In dry red wine In sherry material In port days 1 3 grams per 100 cc 0.337 .250 .200 .168 .144 1-34 grams per 100 cc 0.320 .294 .257 .219 .184 .175 0.157 grams per 100 cc 0.207 .182 .153 .140 .119 104 .085 0.041 grams per 100 cc 0.234 .224 .212 6 7 12 14 28 .177 .153 143 56 76 >31 0.108 Source of data: Marsh, G. L., and M. A. Joslyn. Precipitation rate of cream of tartar from wine. Industrial and Engineering Chemistry, industrial edition. 27:1252-56. 1935. a storage tank. A predetermined amount of bentonite is added (see p. 132) and the wine allowed to cool and become clear. The wine is then fil- tered, after which it is refrigerated to below 20° and stored cold for about a week to a month and then filtered again. The early heating tends to destroy the protective colloids which prevent adequate fining. The addition of tannin at the rate of 1 pound per 1,000 gallons stabilizes the white wines against later clouding and may even improve color retention of red wines lacking in tannin. Removal of excess metallic impurities or prevention of formation of metallic hazes by adjustment of acidity of the wine or by other means is also useful in stabilizing wines. Removal of Cream of Tartar. — As may be seen from the data on cream- of-tartar content given in table 31, the rate of cream-of-tartar deposition 142 Brown, E. M. Stabilization of Wines. Wines and Vines 17(8) : 12-13. 1936. 128 University op California — Experiment Station during refrigeration is slower in dessert wines than in table wines, even though at their higher alcohol and extract content the solubility of cream of tartar would be expected to be lower than and the degree of super saturation to be greater than for table wine. The rate of precipita- tion of cream of tartar from the dry red wine is slower than from the dry white wine. The same difference between white and red wines is shown in the behavior of sherry and port, the rate of precipitation of cream of tartar from the port being slower, largely owing to an initial lag, than from the sherry material. It is surprising that the presence of appreciable tannin and coloring matter in dry red wine and in the port and their higher sugar and extract content would result in a lower rate of cream-of-tartar separation than occurred in dry white wine and sherry material, respectively. No significant difference in the salt and acid content between the red and white wines is to be expected ; such differences in the amounts of these and other constituents as are found in red wines as compared with white ones would tend to decrease the solubility of cream of tartar in the former. It might seem that the presence of tannin and coloring matter in the red wine may have interfered with either the formation of nuclei or the growth of crystals of cream of tartar, since the higher sugar and extract content would have the opposite effect. That the coloring matter does not have a simple retarding effect on the rate of cream-of-tartar separation — that, on the contrary, it may actually increase the rate of tartrate pre- cipitation — was shown in an interesting experience with a California port wine. About 8 liters of port had been stored at room temperature in a 20- liter carboy for about 8 months. This wine was found to be extremely cloudy and to have precipitated most of its coloring matter and a por- tion of its cream of tartar. On warming the wine at 140° F for an hour or two, the cream of tartar redissolved but the precipitated coloring matter did not, being apparently converted to an insoluble oxidized form. The surprising finding was that no cream of tartar separated from a sample of the wine when frozen under conditions such that an appreciable amount of cream of tartar precipitated in the original port. After removal of readily oxidizable coloring material, the rate of pre- cipitation of cream of tartar and the amount precipitated were both markedly less than in the original sample, even though the cream-of- tartar content was the same. The chief changes in composition of wine during refrigeration were decreases in specific gravity, in extract, and in ash which accompanied the removal of cream of tartar and were probably caused by it. A slight increase in pH noted was also probably caused by it. In filtered new white wines the precipitate was practically Bul. 651] Commercial Production of Dessert Wines 129 pure cream of tartar, but in red wine, coloring matter and tannin also are precipitated, particularly on freezing. Little or no precipitation of coloring matter occurs during storage at 32° F or above in the absence of oxidation. FILTRATION Pumping or allowing wine to flow through some media of small pore size — that is, nitration — is one of the oldest and commonest methods of removing particles from wine. For dessert wines it is also one of the most satisfactory. The process is relatively rapid, does not require un- due technical skill, and the results are immediate. In addition, it is a very adaptable method of clarification. It may be used for very cloudy young wines containing large amounts of yeast and colloidal material, as well as for nearly brilliant wines ready to be shipped. Types. — There are two general types of filter processes adapted for use with wines : those which remove particles by adsorption and those which remove particles by virtue of the small pore size of the filtering surface. Filtration by adsorption is utilized in the cellulose type of filters, although there is also a certain amount of filtration due to small pore size. Filtration through material of small pore size is more common and is found in candle filters, in asbestos pad or mat filters, in filters using filter aids, and the like. In general, the latter types are used in California for cloudy young wines and even for polishing filtration of finished wines. The adsorption type of filter is mainly applicable to liquids that have only a small amount of cloudiness, although sometimes fairly cloudy wines are used. The pulp must be thoroughly sterilized before re-use. For wines very much charged with foreign material, a filter with a very large surface is needed for efficient operation. This large surface is achieved by a very large area of filtering surface and also by mixing various filter aids in the wine during filtration. The purpose of these filter aids is to improve the rate of filtration by continually forming new and more rigid filter surfaces. The slimy or colloidal substances are thus unable to collect and form a poor filtering surface. Diatomaceous earths are the most common filter aids. 143 The several types of commercial filters have been discussed in Bulletin 639 (p. 94^-97). The metal portions of all types of filters should be con- structed of corrosion-resistant materials. Filter presses are the most popular and generally useful type. In them canvas sheets are precoated with diatomaceous-earth filter aid and the wine is continually mixed with the filter aid to maintain the rate of flow. The canvas sheets must 143 Calvert, R. Diatomaceous earth. 251 p. The Chemical Catalog Co., New York, N. Y. 1930. 130 University of California — Experiment Station be thoroughly sterilized before re-use. These presses can be purchased in a variety of sizes. Leaf -type filters in which hollow perforated screens are coated with a filter aid are also used. The wine, mixed with filter aid, is forced through the filter surface and screen. At the finish of a run, by a suitable arrangement of the pumps, the direction of flow may be reversed and the coating removed from the screens. These filters are also available in a variety of sizes. The hollow screens are rather expensive and are somewhat difficult to clean. The polishing filters have a small pore size : if used for cloudy liquids they will give only a small rate of flow, and much time w T ill be lost in changing the pads or washing the candle filters; hence they should only be used for relatively clear wines. Asbestos- or fiber-pad filters are the most common. The pads must be thoroughly washed with water before use or they will give an undesirable filter taste to the wine. It is also advisable to wash the pads with a 1 per cent solution of citric or tartaric; acid before use to remove soluble metallic impurities. These pads are rather expensive and are somewhat difficult to wash and sterilize for re-use. They must, therefore, usually be discarded after the filtration. Recently, new filter-pad materials made of glass, rubber, wire, and resin have been introduced. 1 " The cloths made of vinyl resin yarns are resis- tant to acids, though not to heat, and may prove useful for wines. Some of the rubber-pad materials may also prove useful, if they communicate no taste to the wine. Candle filters made of porcelain or carbon are occasionally used in wineries. In some cases the filter is large and as many as 20 filters may be used. For efficient use, however, nearly brilliant wine must be used. In operating any type of filter, some test of the efficacy of the filtration should be made. Visual inspection of the original sample and the filtered wine is sometimes useful, but more adequate control is desirable. Modern industrial plants usually test the brightness of the material being- filtered at successive stages of the filtration. Any convenient nephel- ometer system, such as a photoelectric colorimeter, may be used for this purpose. Wines sometimes foam during routine winery handling, particularly during filtering or bottling. Removal of the foam during fermentation will help reduce f oaminess in wines. New wines foam much more than old. FINING The use of various agents, such as milk, egg white, and blood, for facilitating the rate of sedimentation of suspended material in wine 144 Van Antwerpen, F. J. Filter media. Industrial and Engineering Chemistry, industrial edition 32(12) : 1580-84. 1940. Bul. 651 J Commercial Production of Dessert Wines 131 is very ancient. In modern times these have been largely replaced with agents such as gelatin, albumin, isinglass, casein, Spanish clays, kaolin, and bentonite, as well as various European proprietary products, such as "Saf" and "Coagol." For California dessert wines, gelatin and ben- tonite are the only fining materials that are commonly being used for clarification. Gelatin. — In the use of gelatin, there is a coagulation of the fining agent as soon as it is added to the wine. This is due to the chemical com- bination of gelatin and tannin and also to the influence of the metals of the wine on the colloid. Coagulation is particularly apt to occur if the wine has been aerated prior to or during fining, and if ferric ion is pres- ent. It is sometimes noted that wines after heating do not readily fine with gelatin. Ribereau-Gayon and Peynaud 145 found that this failure of the gelatin to settle was due to the protective colloids present in the wine and that the effect of these protective colloids is considerably increased by heating. Only the smallest amount of gelatin that will clarify the wine should be used. In general, the higher the tannin content and the lower the temperature, the less the amount of gelatin needed. In order to prevent overfining, small-scale laboratory tests should always be conducted with various gelatin-tannin combinations before using them in the winery. The laboratory test must be conducted at the temperature of the winery, if comparable results are to be obtained. Such tests are particu- larly necessary for white dessert wines, such as Angelica, which are very low in tannin. Indeed, it is questionable if gelatin should ever be used for these wines under most California dessert-winery conditions. The high storage temperatures, the very large containers, and the lack of aeration in such containers make the use of gelatin hazardous for white dessert wines. In any event, it should never be used without pre- liminary trials. It has also been suggested that the difference in temper- ature between the top and bottom of very large tanks sets up currents in the wine which continually keep the coagulated agent in suspension or at least prevent perfect settling. Whenever fining agents are used, they should be mixed with the whole mass of wine immediately after their introduction. Bentonite. — Bentonite, a clay, is commonly used in California for fining dessert wines. Clay materials that are somewhat similar chem- ically and physically, called kaolin and Spanish earth, have been used for many years in Europe for wines which are difficult to clarify, and particularly for the sweet wines of Spain and Portugal, according to . 145 Kibereau-Gayon, J., and E. Peynaud. Etudes sur le collage des vins. Eevue de Viticulture 81:5-11, 37-44, 53-59, 117-24, 165-71, 201-5, 310-15, 341-46, 361-65, 389-97, 405-11, 1934; 82:8-13, 1935. 132 University of California — Experiment Station Sannino. 140 Kaolin is a clay of well-known composition, but the proper- ties of the Spanish earths (such as those of Lebrija) are not sufficiently well known, and the analysis published by Sannino shows considerable variation, with some carbonate being present. Miiller 147 also reports that Spanish earth is an impure fine clay (mainly aluminum silicate) and that it is used for the fining of wines of high specific gravity or, in Germany, in conjunction with gelatin for fining wines of highly muci- laginous nature. The amounts recommended, 100 to 400 parts per million, are similar to those recommended in California for bentonite. Ribereau-Gayon and Peynaud recommend kaolin for wines which have been overfined with gelatin. According to Dana and Ford, 148 bentonite occurs widely distributed in western United States and Canada, is derived from the alteration of volcanic ash or tuff, and is largely composed of montmorillonite. Mont- morillonite is one of the kaolin group of silicates, possibly having the formula (Mg, Ca) • A1 2 3 • 5 Si0 2 • n H 2 0, with n varying from 5 to 7. The bentonite commonly used for California wines comes mainly from Wyoming. According to Miiller, the absorption ability of kaolin clays for the color, aroma, and taste components of wines depends on the fineness of the clay and on the nature of the particular clay used. Ben- tonite has the advantage of settling quickly and of not causing cloud- iness when used in excessive amounts. It should not, however, be used in amounts exceeding 700 parts per million. 149 The earthy taste sometimes found in wines treated with clays can be removed, according to Ribereau-Gayon, 150 by the use of a small amount of activated carbon. Other Fining Agents. — Casein is sometimes used for fining dessert wines, particularly in conjunction with bentonite or for wines that are high in iron. Laboratory tests should be conducted on the wines in order to determine the proper amounts to use. Only good-quality casein should be used. Fresh defibrinated blood of cows has been used but is not considered desirable because of the possible health hazards and danger of unpleasant publicity. Fining versus Filtration. — Both fining and filtration have their ad- 140 Sannino, F. A. Trattato completo di enologia. 2d ed. Vol. II. 368 p. (See espe- cially p. 95-101.) Vincenzo Bona, Torino, Italy. 1920. 147 Miiller, K. Weinbau-Lexikon, 1015 p. Paul Parey, Berlin, Germany. 1930. 148 Dana, E. S., and W. E. Ford. A textbook of mineralogy. 4th ed. 851 p. John Wiley and Sons, New York, N. Y. 1932. 140 Say well, L. G. Clarification of wine. Industrial and Engineering Chemistry, in- dustrial edition 26:981-82. 1934. (See also: California Wine Eeview 3(1):14-15, 24. 1935.) ^° Eibereau-Gayon, J. Clarification des vins. Eevue de Viticulture 82:367-68. 1935. Bul. 651] Commercial Production of Dessert Wines 133 vocates, and in dessert-wine making both are usually necessary. Filters are more expensive than fining agents but require little special technical training to operate. Fining agents, on the other hand, must be used with care, particularly as to the amounts and the temperatures at which they are applied. The fining agents are generally more efficacious in producing a permanently brilliant wine when properly used, but fil- tration is a more rapid process and can be used at any time during the year. The aeration resulting from filtration is usually beneficial in dessert wines in contrast to the effect on table wines, although undue filtration of delicate, aged dessert wines may be harmful. While filtra- tion is undoubtedly of very great importance, the use of the various fining agents should be more generally investigated by the wineries producing dessert wines. CENTRIFUGING Centrifuging is more common for California dessert wines than for table wines. The wine sometimes benefits from the aeration received, and some colloidal material, not easily filtered out, may be removed. The equipment is somewhat expensive, however, and centrifuging is not as generally used in the winery as filtration. PASTEURIZATION Table wines are pasteurized to destroy microorganisms which are capable of producing disease in them, but dessert wines are seldom pasteurized for this purpose except for wines that are very high in mi- crobiological content, for those that are low in alcohol content, or for those that are most subject to contamination. They may be heated, however, for the purpose of assisting the clarification. Some forms of colloidal material are coagulated by heat treatment, but others may be permanently dispersed by pasteurization. For this reason it is better to heat only wines that are free of suspended material. The pasteurization, by its precipitation of some material, favorably aids the permanent clarification of certain dessert wines. A laboratory test should be made to determine if pasteurization will cause the pre- cipitation of organic material. A heat exchanger in which flash pasteuri- zation is possible is the usual means of accomplishing pasteurization. It should be constructed of corrosion-resistant metal throughout. Copper particularly should be avoided. A short period, a high temperature, and a continuous unit in which the wine is rapidly cooled is considered better than a long period, or a lower temperature, or a discontinuous type of heating. Heating to 180° F for 1 minute is the usual California practice. 134 University of California — Experiment Station PREPARATION FOR MARKETING 151 BLENDING Blending is one of the most important operations in the winery pro- ducing dessert wines. When properly conducted it is of great value to the winery; but when it is improperly carried out, good wines may be spoiled. Purpose. — The purposes of blending are: (a) to maintain standard types, ( b ) to add to the character and bouquet of aged wines the fresh- ness of young wines, and vice versa, (c) to balance the composition of wines, and (d) to maintain a constant supply of aging wine of a certain character. Far too little attention has been paid to the constructive values of blending in many post-Repeal plants. The very large tanks and the lack of aged stocks have, of course, greatly restricted the possibilities of blending or even made it impossible. Owing to this lack of suitable blends, some distributors have been doing their own blending. Many of the European dessert wines owe much of their excellence to skillful handling and especially to blending after fermentation. Euro- pean practice therefore includes partial fortification at fermenting rooms adjacent to the vineyard prior to, during, or immediately after fermentation and then shipment to the large central storage cellars. The wines arrive at the storage cellars with greatly varying degrees of alcohol and sugar, and often, in the case of ports particularly, are made from different mixtures of grapes. They are, however, immediately blended and refortified to make a few uniform products. They are then stored in butts or pipes of oak or in other small oak containers. Some- times these are so arranged as to provide for progressive fractional blending during maturation (solera system, see p. 100). The wines withdrawn for shipment are then further blended to suit the trade. These procedures allow the marketing of standardized types of sweet wines from year to year, even though the wines of any given year may differ appreciably from those of other years. Blending wines as they age, as in the solera system in Spain, is, how- ever, an expensive system, as very large stocks must be maintained, and it has been very little practiced in this state. The system not only demands large stocks, but it is expensive to maintain owing to the large number of casks to be filled and refilled. This system or a modifica- tion of it might, however, prove useful in California for dessert wines even with large-sized containers. To set up such a blending system, a cask or tank of the best wine should be set aside. About every six months 151 General references on this subject in addition to those given in specific footnotes in the section are listed on page 176. Bul. 651] Commercial Production of Dessert Wines 135 or year after this wine matures about a fourth of the cask or tank should be drawn off. The secret of the success of the system lies in finding a very similar wine to refill the original container. This second container is in turn filled up with some younger but similar wine, and so on, so that eventually there may be from four to ten stages. The constant blending and the fact that only from a fourth to a half of any single cask is drawn off at one time leads to the production of a standard type which varies only slightly over a period of years. The greatest difficulty, however, is to get wines whose character and composition closely re- semble those of the original type of wine for the filling of subsequent casks and not to draw on the casks too frequently or for too great a per- centage of the wine. The most important functions of blending dessert wines are to obtain standard types and maintain an identical character for them over long periods of time. The continual change in the character of the wine bottled as a certain type and brand is undesirable for it may lead to unfavorable consumer prejudice. This is particularly true for the more experienced clients who become familiar with a certain company's California port under a given label and who can, and will, detect changes in the wines of different seasons unless careful blending is practiced. The types and brands which a company may prepare for sale will depend on the stocks available as well as on the consumer's demands. Accurate records of the composition, particularly of the alcohol, total acid, sugar, and hue and depth of color, should be maintained for each type. Bottled wines should also be kept for future organoleptic (smelling, tasting, and visual) comparison. In blending, a small proportion of aged wine, particularly that aged in small oak casks, will frequently add just the desired cachet to the wine. Some aged wines, on the other hand, lack the light color of a young Angelica or the fruity muscat flavor of a young muscatel. Blend- ing in of younger wine is a means of improving such wines. Many wines lack a single ingredient or contain an excess of some ingredient. Proper blending will correct for these deficiencies or excesses — sugar and color are the commonest. The winery should always prepare some wine of higher and lower sugar and alcohol content than the normal to correct for such accidental discrepancies or to assist in preparing wines of any desired composition for a particular market. The alcohol content is occasionally adjusted by refortification. Refortification, however, can be done only in certain special cases and under government regulations. The sugar content of wines may similarly be adjusted either by the addition of sweetening agents prior to fortification or by blending wines of higher sugar content with those of lower sugar content. Grape 136 University op California — Experiment Station concentrate is the only sweetening agent that can be used in ameliorat- ing California wines. This may be used in either or both of two ways. It may be added to the wine before fortification to raise its sugar con- tent to the desired point, or a medium-sweet concentrate containing y 2 of 1 per cent of alcohol by volume may be fortified to an alcoholic content of up to 24 per cent and blended with the newly fortified wine. The latter procedure is preferred because it yields wines that are smoother in flavor. 152 Sweetened blending wines obtained by adding must or concentrate to wine and then fortifying it with brandy are used also. The acid content of the wines should be adjusted before aging and preferably before fermentation. (See p. 44 of this bulletin and p. 35-37 of Bulletin 639.) Tannin may be added not only for stabilizing wines but also to obtain wines of better balance. Blending out off -flavors or high volatile acidity is a dangerous prac- tice. It is much better to dispose of wines having these defects at a minimum price rather than to spoil or impair the quality of good wine by blending and then receive only a low price for the whole blend. Technique. — Successful blending is not easy. It requires a keen memory for small differences in odor and flavor and a clear and fixed initial idea of the type of wine that is wanted. As mentioned, compara- tive samples of previous blends are useful. The cost per gallon of each wine in the cellar should be known. Only those wines which are available (because of price, quantity, or other considerations) should be tasted. The previous tasting records and a copy of a recent analysis of the wine should also be available. Trial blends are conveniently made in large, graduate cylinders, in calibrated tasting glasses, or with pipettes. After deciding on the proper proportions for a blend, a new blend with the desired percentages is made. If this test is still favorable, a trial blend of 10 to 50 gallons should be made, and stability tests conducted in the wood and in bottles at various temperatures. After making blends, the blended wines should be allowed to age in the cask or tank from three weeks to six months before bottling, to permit any cloudiness developed by the blending to be detected and removed. In blending wines to obtain a desired standard, the relative quan- tities of the wines to be used may be calculated if the composition of the components of the blend is known. If only one constituent — sugar, 152 A ruling of the United States Bureau of Internal Revenue requires that con- centrate per se may be added to wine only prior to fortification. Grape products test- ing up to about 40.0° Balling after fortification, however, may be used in blending. Such products are classified as wine for blending purposes only. Bul. 651] Commercial Production of Dessert Wines 137 alcohol, or acid — is changed while the others remain constant, the blending formula is simple and readily applied. If more than one con- stituent is changed, then by adjusting first the alcohol content, then the sugar content, and so forth, the actual case may be made to approximate the theoretical. (More complicated formulas containing more than one unknown may of course be used.) If A and B are the weights of two ingredients to be mixed to produce a weight of (A-\-B) of the mixture, and a and b are the percentages by weight in the two ingredients of some component common to both of them and if m is the percentage of the same component in the final mix- ture, then it can be shown that the proportion by weight in which the ingredients must be mixed to yield a product of the desired composition A m-b is: — = B a-m If a, b, and m are percentages by volume — of alcohol, for example — and if the changes in volume that occur on mixing are neglected, then this is also the proportion by volume in which the two wines are to be mixed. For example, if the alcoholic content of wine A is 21.0 per cent, and that of wine B 19.5 per cent, and a blend containing 20.0 per cent is desired, then the proportion by volume in which A and B are to be . , . . . A 20-19.5 0.5 1 mixed is very closely: £ = 2l=2O0 = L0 = 2 ' If a, b, and m are percentages by weight — of sugar, for example — and if the changes in volume that occur on mixing are neglected, and the specific gravities of the components assumed to be close enough to be the same, then the ratio by volume in which the two wines are to be mixed also is given by the above relation. For example, if the re- ducing-sugar content of a sherry A is 4.0 per cent and of a sherry B is 0.5 per cent, and a blend containing 2.0 per cent is desired, then the pro- portion by volume in which A and B are to be mixed is approximately A = 2.0-0.5 1.5 _ J_ B " " 4.0-2.0 = 2.0 — 4 * This algebraic formula may be more easily depicted by the Pearson square as follows : wine A a\ \m-b wine B ¥ >a-m If a fairly dry muscat wine testing 2° Balling is blended with a for- tified unfermented muscat juice testing 20° Balling, the proportion in 138 University of California — Experiment Station which the two must be mixed to obtain a muscat testing 8° is approxi- mately 20 6 or 6 parts of juice to 12 parts of the dry muscat wine, or 1:2. The Balling degree, however, does not indicate the true solids content of a wine, as discussed in the section on vinification (p. 45) . If the wines in the example given all contained 20.0 per cent alcohol by volume, then the sugar content on an alcohol-free basis would correspond to 8.09 per cent and 25.30 per cent for the two components of the blend and 13.93 per cent for the mixture. The proportion in which the components should be mixed would be 5.84 parts of the juice to 11.37 parts of dry wine, or 1 :1.92 instead of 1 :2. If the alcoholic contents of the components of the blend were different, the error introduced would be considerably greater. TASTING As outlined in the previous section, intelligent blending can be ac- complished at the present time only on the basis of a thorough exami- nation by smelling and tasting. Tasting is particularly important with dessert wines in which, by blending, the maintenance of a standard from year to year is attempted. The tasting should be done in a special room which has sufficient light and is free from winery odors. A supply of cool and of warm, clean, soft water should be provided. The best glasses for tasting are the tulip-shaped thin-walled types (see Bulletin 639, p. 105). After use they should be thoroughly washed so that no odor or flavor carries over to the next wine. When samples are being taken directly from containers, the taster should assure himself that the wine thief used for taking the samples is clean. The thief should be sufficiently long to secure the sample well down in the container. In examining the color of dessert wines, the standards commonly ac- cepted by the wine industry and particularly those used by the par- ticular winery must be considered. Visual examination and comparison with previous samples is valuable, if only an approximate duplication is desired. For more complete color specification, see page 161. A brownish tinge is not an infallible indication of age in California, where dessert wines are frequently heated. A tawny color accompanied by a clean, noncaramelized, aged aroma is desirable in ports. Some producers secure a tawny color by heating, but the aroma and flavor of such wines is not the same nor as desirable as that of the normally aged wines. Bul. 651] Commercial Production of Dessert Wines 139 California pale sherrys should be produced naturally from light - colored wines. The regular sherry has a medium-amber color. Muscatel and Angelica should have a light-gold color. California Tokay should be a medium amber with a tinge of red. Excessive cloudiness in dessert wines is objectionable, but the pres- ence of small amounts of precipitate due to aging are entirely normal in dessert wines which are aged in bottles and may be found after a period of time in the bottle even from wines which have been "finished" before bottling by the most thorough treatments. The aroma and bouquet of dessert wines are particularly important. The quality of the brandy used in fortifying should first be established. The presence of too high an amount of "heads" (aldehydes) or "tails" (higher alcohols) can be detected by smelling. The amount of grape aroma present, particularly in muscatels, can also be determined, as well as that of excessive amounts of overripe grapes, concentrate, or boiled-down musts. Overly enthusiastic use of filter aids, charcoal, oak chips, heat, and other treatments may also be recognized in some Cali- fornia wines by the critical taster. An effort should be made, where blending tests are to be attempted to match a previous wine, first to identify the particular characteristic aroma and flavor of the standard which it is desired to duplicate. This includes a good memory for the original condition of the standard if the new blend is to be aged. Even if the blend is not to be aged, the merits and defects of the standard should be known in order to make the blend with the least possible amount of trial and error. Because of the sugar present in most dessert wines and their high alcohol content, the actual tasting of these wines is cloying and dehy- drating to the palate. Only a small amount should actually be introduced into the mouth, and the wine should all be expectorated as soon as pos- sible. Frequent washings of the mouth with water will help prolong the tasting period, but, at best, it cannot be continued very long. Great attention is therefore directed to the appearance and aroma of the wine by careful tasters, who wish to prolong the period of effective tasting by reducing the amount of actual gustatory examination. BOTTLING Much of California dessert wine is sold in bulk but the best is bottled. Wines are commonly sold by volume ; but the volume of wine changes with temperature, expanding as the temperature increases and con- tracting as the temperature decreases. Water changes in volume by ± 0.021 per cent of its volume at room temperature (68° F) for each 18° F change in temperature, alcohol changes by ± 1.05 per cent. The 140 University of California — Experiment Station expansion of a given weight of alcohol is 3% times that of same weight of water. The coefficient of expansion of sugar solutions is lower than that of water; and that of dessert wine is intermediate, depending largely on alcohol and sugar content. It is sufficient, however, to cause some unfounded complaints of shortage when wine is received at tem- peratures lower than those at which it was measured and may cause breakage of containers that are filled too full. On freezing, wine, like water, expands in volume, and bottled wines should be protected against extreme fluctuations in temperature in transit or storage. Because of the large cubical coefficient of expansion of wine, care should be taken in barreling to fill the wine barrels at a standard temperature, usually 60° F, or if the temperature is over 60° to fill to above 50 gallons so that the wine will not contract to below 50 gallons on cooling. Wine barrels are usually made to 54 gallons' capacity. Shippers should state plainly to buyers that a standard barrel is 50 gallons but that additional space is provided for expansion; that headspace in a barrel does not indicate that the buyer is not getting 50 gallons of wine. In shipping large con- tainers of wine, headspace to allow for expansion must be provided. The container may be protected against breakage by providing it with a vent hole or low-pressure safety valve. Preparation for Bottling. — The wine maker must be sure that the wine is properly blended and clarified before bottling. But the character of the wine should not be harmed by excessive treatment. The fruity varietal flavor and light color of muscatel are its chief attractions, and these should not be lost by undue aging or treatments prior to bottling. Most California dessert wines are filtered through a pad filter just prior to bottling. These filters should be regularly cleaned and their perform- ance checked frequently during bottling. If necessary, amounts of sulfur dioxide up to 50 or 75 parts per million may be added to control unde- sirable microorganisms. (See p. 146.) Bottles. — There is far too wide a variation in the size and shape of bottles used for California dessert wines. In the pre-Prohibition era, quart bottles, usually brown or green in color, were used for California wines. A somewhat smaller bottle is now ordinarily used. The industry should standardize on some one size. Many odd and some fantastic- shaped bottles are now found on the market. Although some of these may prove amusing to a certain clientele, they are not designed to conform to custom or to provide a bottle which may be laid on the side for aging. Well-proportioned and desirable bottles for dessert wines are shown in figure 13. The general use of mechanical fillers makes it necessary that the bottles should be of a good strength in order to stand the movement Bul. 651] Commercial Production of Dessert Wines 141 through the bottling machine without breaking. A few bottles from each lot purchased should be examined by means of polarized light to deter- mine the relative freedom of the bottles from defects. If too many bottles are of uneven thickness, nonuniform color, or contain air bubbles, the lot should be rejected. Also, if breakage in the bottling line is high, the bottle manufacturer should be warned. Some difficulty with certain types of bottles has been experienced because of the presence of a high aluminum content in the glass. If this type of cloudiness or precipitation is suspected, the sediment should be examined chemically. Fig. 13. — Bottles used for dessert wines, illustrating shape and type of closure. The bottles for sweet wines should be washed before using. Used bottles are not recommended. The bottles may be washed with some detergent — soda ash, trisodium phosphate, and metaphosphate solu- tions are commonly used — rinsed with clean, sterile water, and allowed to dry before filling. The cleaned bottles should not be allowed to remain in the winery unfilled, lest they collect dust or become contaminated. Bottling. — Formerly filling was done largely by hand directly from the cask. In small wineries this method is still used. For the moderate- sized winery, fillers which are semiautomatic and handle six or more bottles at a time may be used. The larger wineries should use the com- pletely automatic fillers. This latter type of filler is expensive; but it requires little attention, fills each bottle with the same amount of liquid, and can be obtained in a variety of styles and sizes. The spouts should reach nearly to the bottom of the bottle in order to prevent excess foam- ing and oxidation. The bottling room should be light, easily cleaned, free of sources of contamination, and well ventilated. The fillers must be kept clean dur- 142 University of California — Experiment Station ing use, and before and after operating they should be washed and sterilized. A little of the wine to be bottled may also be rinsed through the machine before starting. This wine must be discarded. All equip- ment with which the wine comes in contact, from the polishing filter to the bottle, should be of corrosion-resistant material. Considerable econ- omy in operating the bottling room can be effected in the large plant by mechanizing all movements of the bottles and arranging the equipment properly. (See fig. 10, p. 70.) Dessert wines susceptible to spoilage by lactic-acid bacteria should be either filled hot into the bottle or pasteurized after bottling. The bottles should be filled, whether by hand or by mechanical filler, well into the neck of the bottle. The space remaining in the neck should be uniform, and according to Pedersen and co-workers 153 it is best even with fortified sweet wines such as port, to fill the bottle almost full in order to prevent undesirable changes in storage. Closures. — Screw caps alone or in conjunction with short corks are more widely used for dessert wines in California than straight corks. The cap should seat properly and seal the bottle completely. Certain types of metal closures are made so that a seal must be broken before opening. These assure the buyer that the wine has not been tampered with before purchasing. Other types of closure are made in one piece with the metal capsule. These also have a tamperproof seal in certain cases. Although the tamperproof features are good, the important points, for any type of closure, are that the closure hold the wine and not leak nor communicate any foreign taste to the wine. If the dessert wine is to be aged in the bottle, straight corks are desirable. These corks should be of good size so that leakage does not occur. Sometimes corks with a wooden top are used so that the cork may be removed and returned to the bottle. These are appropriate for dessert wines, since the wine will not spoil easily in a half -full bottle. Capsuling and Labeling. — The final operations before shipment are placing a metal foil or cellulose cap over the closure and adding the proper labels. While all of these operations were once done by hand, machinery is now available for most of them. Automatic labeling appa- ratus is particularly useful. The names used for California dessert wines are outlined on pages 21-33. The information required by law on the label varies from state to state and is frequently changed. The Wine Institute or the Alcohol Tax Unit of the United States Bureau of In- ternal Revenue should therefore be consulted before any labels are printed. General information such as : type of wine, contents of bottle, ins Pedorsen, C. S., H. E. Goresline, A. L. Carl, and E. A. Beavens. Keeping quality of bottled wines. Industrial and Engineering Chemistry, industrial edition 33:304-7. 1941. Bul. 651] Commercial Production op Dessert Wines 143 alcohol content, and bottler's name, are required. Some statements are prohibited (for example, misleading statements of quality). The size of type which may be used is specified for certain information printed on the label. The design on the label should be simple and dignified. BACTERIAL DISEASES AND OTHER DISORDERS OF DESSERT WINES 354 The higher alcoholic content of fortified dessert wines renders them less susceptible to many of the bacterial diseases, and their higher pH value to the disorders, to which the table wines are subject (Bulletin 639, p. 120-27.) Still the dessert wines are subject to bacterial diseases during fermentation and prior to fortification and may even undergo spoilage by alcohol-tolerant microorganisms after fortification. Oxida- tion and reduction of the wines during aging may bring about darkening and discoloration, or formation of turbidities and deposits similar to those formed in wines of lower alcohol content. The higher sugar con- tent, the higher alcohol content, and the lower acidity modify the nature and extent of this nonbacterial spoilage. MICROBIOLOGICAL SPOILAGE During Fermentation. — The most serious spoilage observed during fermentation is the rapid acetification of musts fermented at high tem- peratures without the use of sulfur dioxide or pure yeast. Cruess and Quaccia 155 observed the production of as much as 0.4 grams of acetic acid per 100 cc of muscat must in 48 hours under such conditions. Later observations by Cruess 156 on similar abnormal fermentations which oc- curred during the 1936 season, another hot season like that in 1934, showed that such rapid acetification could be prevented by use of sulfur dioxide, pure yeast starters, and proper cooling during fermentation. While muscat must was most susceptible to this type of spoilage, it was observed also in Flame Tokay and Carignane musts. Cruess and Quaccia in 1934 believed that the causative agent was a yeast capable of form- ing abnormally high amounts of volatile acid, but later (1937) Cruess reported that microscopic examination showed the presence of long rod-shaped bacteria believed to be lactic-acid bacteria. Vaughn, 157 however, found the causative organisms to be several unusual strains 154 General references on this subject in addition to specific footnotes given in the section are listed on p. 176-77. 155 Cruess, W. V., and L. Quaccia. Observations in spoiling of must by yeasts. Fruit Products Journal 4:109. 1934. 158 Cruess, W. V. Observations of '36 season on volatile acid formation in muscat fermentations. Fruit Products Journal 16:198-200, 215, 219. 1937. 157 Vaughn, Reese H. Some effects of association and competition on Acetohacter. Journal of Bacteriology 36:357-67. 1938. 144 University of California — Experiment Station of Acetobactcr. Grapes heavily infested with these acetic acid bacteria, when crushed and fermented at high temperature, undergo an incipient alcoholic fermentation at first because these bacteria do not attack sugar as rapidly as yeast does. As soon as small quantities of alcohol are pro- duced, however, the alcohol is rapidly converted into acetic acid by these bacteria. The accumulation of acetic acid and the prevailing high tem- perature retard or prevent the activities of yeast, and the must sticks. Lactic acid bacteria may be involved in this spoilage also since in asso- ciation with acetic acid bacteria they too produce volatile acids as well as fixed acid. Such a must is "mousy" in flavor and should be discarded because it cannot be refermented and made into a satisfactory wine even for distilling material. To control this spoilage Cruess recommends the addition of not less than 100 parts per million of sulfur dioxide (S 1 /^ ounces liquid sulfur dioxide or 7 ounces of potassium metabisulfite per ton of grapes crushed). Cooling during fermentation is necessary also. (See p. 42.) The acetic acid bacteria are extremely sensitive to sulfur dioxide. After Fortification. — Fortified dessert wines, because of their high alcohol content, are relatively resistant to attack by the common rod- shaped lactic acid bacteria which are so often found in table wines. Spoil- age of dessert wine by alcohol-tolerant lactic acid bacteria has been reported to occur by several investigators in recent years. Fevrier 1 '' 8 reported a bacterial disease in South African wines containing 18 per cent alcohol, and Niehaus 159 reported "mannitic" bacteria in wines from the same locality. Otani 160 described four new species of lactobacilli capa- ble of growing in sake containing from 22 to 26 per cent alcohol. For- nachon 161 isolated rod-shaped lactic acid bacteria from fortified Austral- ian wines. Although several instances of the occurrence of similar lactic acid bacteria in California fortified wine have been observed, 162 they have not been investigated in detail. The most common and serious disease of fortified California wines is caused by an organism which, because of its resemblance under the micsroscope to a tangled mass of wet hair (fig. 14) has been called the "hair bacillus"; it has been termed also "cottony mold," and "Fresno mold." Its occurrence in fortified wines was reported upon by Douglas 158 Fevrier, F. A bacterial disease in wine. Union of South Africa Department of Agriculture Journal 12:120-22. 1926. 159 Niehaus, C. J. G. Mannitic bacteria in South African sweet wines. Farming in South Africa 6:443-44. 1932. 160 Otani, Y. Untersuchungen ueber der Hyochi-bazillen im Sake. Hokkaido Impe- rial University Faculty of Agriculture Journal 34 (pt. 2) : 51-142. 1936. ioi Fornachon, J. C. M. A bacterium causing disease in fortified wine. Australian Journal of Experimental Biology and Medical Science 14:215-22. 1936. 162 Douglas, H. C, and R. H. Vaughn. Unpublished observations. Bul. 651] Commercial Production of Dessert Wines 145 and Cruess in 1936, 163 and a more detailed report made by Douglas and McClung. They describe the spoilage as follows : Spoilage is characterized by an extensive, flocculent, amorphous sediment in the bottom of the container with the supernatant liquid remaining perfectly clear. At no time does the wine, unless shaken severely, become turbid as it does with other bacterial diseases. Chemical examination of such wines reveals ordinarily that there has been but little change in composition. Upon prolonged storage at room tempera- 1 *> * ' ; -.'.,•' . -. life : ■■■-.«. .**«■■:■'..;.■-'',■-: ;;:d:\'t... , ■ :-.,,. .. ■■'. , . ' *'::. ■■ ■ ; . I '' ' , „ ■„ '''" I • - pmm J m . ', <''kC^y.^- t .-'- ,„ ■ : ' ■ ' "" - -'' .: -f /- Si „>'* ' ; ^9 Wm '.. ■■■■:.:: ■,■. Fig. 14. — Typical unstained microscopic appearance of the organism causing flocculent, amorphous sediment in dessert wines. Magnification approximately 1,000. (Preparation by courtesy of H. C. Douglas.) ture, there is an increase in the volatile and fixed acids of the wine, together with a decrease in reducing sugars and the pH. An occasional sample is found in which some gas has been produced. Microscopic examination of the sediment from such wines reveals masses of long, intertwined filaments. Most frequently affected in the samples examined were Muscatel, Angelica, and sherry types of wine, with occasional samples of port, Tokay, and Malaga. We have encountered no report of spoilage of this kind in dry wines. The most frequent complaints in this type of disease are from instances in which spoilage occurs after final bottling. Although not affecting the taste greatly, diseased wine is unsightly and the consumer's reaction is naturally unfavorable. Such a reac- tion may be expected to affect future sales of similar or other types of wines from the same wineries. Since prolonged incubation is often necessary before spoilage takes place, the possibility exists that wine which is apparently sound upon leaving the winery may later develop the growth. The case histories of some of the complaints indicate that this condition prevails in many instances. 163 Douglas, Howard C, and W. V. Cruess. A note on the spoilage of sweet wine. Fruit Products Journal 15:310. 1936. 146 University of California — Experiment Station Although the majority of reports of this type of spoilage have been upon bottled wines, positive cultures have been obtained from storage vats, shipping barrels, and tank cars, indicating that wine is often contaminated in the winery before bottling. The possibility exists, however, that at times sound wine may become diseased through the use of contaminated bottling equipment or other equipment with which the wine may come in contact. 16 * The causal organism, the filamentous lactic acid bacteria, develops very slowly even under favorable environmental conditions. Its alcohol tolerance is high for it will grow in diluted sweet wines containing up to 22 per cent of alcohol, but it is very sensitive to sulfur dioxide, even 60 parts per million being sufficient to check development of the organism. The sugar content is not the limiting factor, for the organism will grow in wines containing from 0.2 to 15.0 per cent sugar. The lower the total acid content of the wine the greater its susceptibility to this organism : a total acidity of 0.5 per cent is sufficient to prevent its growth in sweet wine. The optimum pH for growth lies between 4.1 and 4.3 with growth occurring from pH 3.7 to between 5.0 and 6.0. The optimum pH for growth and for metabolism of carbon compounds for several strains of heterofermentative (non-gas-producing) lactic acid bacteria was found by Fornachon, Douglas, and Vaughn 165 to be pH 5 to 6. The optimum growth temperature was 68° to 77° F; at higher temperatures the incu- bation period is prolonged and in some cases growth is inhibited alto- gether. Recent investigations 166 indicate that the organism produces large < I nan titles of mannite from levulose. The mannite present in infected wines crystallizes out in characteristic rosettes of needlelike pointed crystals when a portion of the wine is allowed to dry. Lactic and acetic acids are produced also but in small amounts; gas production was not observed by Douglas and McClung in artificially inoculated media in contrast to conditions in several original wine samples. To control the spoilage by this organism, sulfur dioxide, pasteuriza- tion, or acidification may be used. The infected wine should be filtered, pasteurized by ordinary winery pasteurization procedures, and sulfited to a concentration of 60 to 100 parts per million of sulfur dioxide in the finished wine. Yeast turbidities have been observed recently in bottled fortified wines. Whether these turbidities are actually due to development of alcohol -tolerant yeast in the wine, to growth of yeast on corks, or to 164 Douglas, H. C, and L. S. McClung. Characteristics of an organism causing spoil- age in fortified sweet wines. Food Eesearch 2:471-75. 1937. 105 Fornachon, John C. M., Howard C. Douglas, and Reese H. Vaughn. The pH re- quirements of some heterofermentative species of Lactobacillus. Journal of Bacteriol- ogy 40:649-55. 1940. 166 Vaughn, Reese H., et al. Unpublished observations. Bul. 651] Commercial Production of Dessert Wines 147 other causes, is not known. Phaff 167 has isolated from such wines viable species of Zygosaccharomyces. In beer, yeast turbidities occur only in well-aerated beer, that is, at a rather high oxidation-reduction potential, rH above 13 ; 188 otherwise the propagation stops and the yeast cells pre- sent precipitate. 160 That a similar relation occurs in wines is known from the fact that addition of sulfur dioxide inhibits yeast turbidities of this type. In using sulfur dioxide to control the microbial spoilage of fortified sweet wines, it is necessary to realize that the rate of loss of the actively antiseptic, free sulfur dioxide is higher in such wines than in dry table wines because of the more rapid combination of sulfur dioxide with sugar, aldehydes, and other substances. The total sulfur dioxide content is thus not a reliable criterion of the potential preservative power. NONBACTERIAL DISORDERS Hazing, clouding, discoloration, and browning of wine and formation of crystalline or amorphous deposits may occur in young or improperly stabilized wines or those which contain excessive quantities of metallic impurities. The crystalline deposits most commonly observed are those of cream of tartar, calcium tartrate, or mannite. The presence of potassium acid tartrate crystals (cream of tartar) indicates either too short an aging period or insufficient refrigeration, the temperature being too high or the time refrigerated too short. Calcium tartrate crystals occur when the wine dissolves excessive quantities of calcium from improperly washed filter pads, from concrete tanks, or from other sources. Occasion- ally wine filtered with a diatomaceous filter aid may contain small diatom particles which will settle out. The presence of mannite crystals indicates that the wine has been subjected to attack by one or more types of lactic acid bacteria. Amorphous Deposits. — Several types of amorphous deposits are en- countered in dessert wines : precipitates of oxidized or decomposed tannins and anthocyanin pigments; precipitates of insoluble caramel; precipitates of excess tannin ; precipitates of colloidal iron and copper 167 Phaff, H. J. Unpublished observations. 188 rH is the negative logarithm of the calculated hydrogen pressure at which the hydrogen electrode would indicate the same oxidation-reduction potential at the same pH as that of the system under investigation. The higher the values of the rH, the greater is the oxidizing intensity of the system. The rH convention assumes that the electrode potential varies with the pH as does the hydrogen electrode, an assump- tion which is not strictly correct. 189 Siebel, F. P., and E. Singruen. Application of oxidation-reduction potential to brewing control. Industrial and Engineering Chemistry, industrial edition 27:1042-45. 1935. Clerck, Jean de. rH and its applications in brewing. Institute of Brewing Journal 31 (n.s.) :407-19. 1934. 148 University of California — Experiment Station salts; and precipitates of coagulated gums, proteins, or other colloid, and other organic constituents that may be thrown out of solution. Gallotannin, obtained principally from gall nuts, is present also in Vitis vinifera grapes, according to Girard and Lindet. 170 This on mild hydrolysis yields glucose and the tannic acid of commerce. Phlobatannins are present also. 171 The gallotannins and the phlobatannins on oxidation yield highly polymerized insoluble brown-colored complex compounds usually referred to as melanogallic acid (in case of gallic acid) or sim- ply as melanins. The phlobatannins, the most widely distributed nat- ural tannins, in acid solution are slowly converted into red or brown amorphous, insoluble materials that have been named "phlobaphenes." (See also p. 126.) It is possible that "phlobaphenes" occasionally deposit in wines, particularly when they are heated. Laborde 172 suggested two courses for the oxidation of tannin : either direct oxidation of the tannin material by oxygen followed by oxidation of alcohol to aldehdye ; or oxidation of alcohols to aldehdyes first followed by precipitation of an aldehyde complex of tannin. Ribereau-Gayon 173 found that tannin is an antioxidant in wine and reduces the rate of oxidation of oxidizable constituents like sulfur dioxide and anthocyanins. The stabilization of red wine against precipitation of coloring matter on oxidation has been known since 1905. It is well recognized in the wine industry that the addition of 1 pound of tannin (as grape tannin) per 1,000 gallons sta- bilizes wine (both white and red) against sedimentation of coloring matter and other oxidizable organic constituents. The precipitation of red coloring matter was at first believed to occur simply as the result of the direct oxidation of anthocyanin pigments. Trillat (cited in footnote 99, p. 90), Laborde, and Joslyn and Comar 174 have presented evidence to show that it is caused primarily by combina- tion with aldehyde either added or formed by oxidation of alcohol. Ribereau-Gayon indicates that anthocyanins undergo progressive de- methoxylation to yield a derivative capable of polymerization and undergo coagulation if necessary even in the absence of oxygen. 170 Girard, A., and Lindet. Sur le phlobaphene du raisin. Societe Chimique de France Bui. 19(3 serie) : 583-84. 1898. 171 For discussion of the natural tannins see : Nierenstein, M. The natural organic tannins. 319 p. J. & A. Churchill Ltd. London, England. 1934. Eussell, Alfred. The natural tannins. Chemical Reviews 17:155-86. 1935. 172 Laborde, J. Recherches sur le viellissement du vin. Revue de Viticulture 48: 225-30, 241-44; 49:38-41, 49-51, 65-69. 1918. 173 Ribereau-Gayon, Jean. Contribution a l'etude des oxydations et reductions dans les vins. 2d ed. 213 p. (See especially p. 52-55, 92-97.) Delmas-^diteur, Bordeaux, France. 1933. 174 Joslyn, M. A., and C. L. Comar. The role of aldehydes in red wine. Industrial and Engineering Chemistry, industrial edition 33:919-28. 1941. Bul. 651] Commercial Production of Dessert Wines 149 Iron Casse. — The formation of colloidal ferric phosphate, which sub- sequently settles out to form a white deposit, has been observed in white dessert wines such as Angelica and muscatel, and particularly in white port. Marsh 175 believes that this occurs, even in dessert wines, only in the range of pH 2.9 to 3.6. California dessert wines usually range from pH 3.7 to 4.1, the bulk being in the range 3.9 to 4.0. Iron casse, however, occurs in dessert wines, particularly when excessive amounts of both iron and copper salts are present. It is encountered frequently in white port. The more heavily decolorized is the white port, the more susceptible is the wine to the formation of a white acid-soluble sediment. This is due to a pickup of phosphate from the charcoal, particularly from bone charcoal. The ferric phosphate casse can usually be controlled by the addition of 1 pound of citric acid per 1,000 gallons of wine. In wines containing large amounts of both iron and copper, ferric phosphate casse may form even in wines above the pH range in which it normally occurs and may not be controlled by addition of citric acid. Copper Casse. — Copper casse has been found to occur in several white dessert wines containing excessive quantities of copper. Copper casse is seldom a serious problem with sherries even when they are baked with copper coils although copper coils may be undesirable for other reasons. (See p. 106.) Apparently the copper dissolved by the sherry material is largely deposited during the usual cellar treatment. Occasionally, however, sherries with excessive copper content occur. Successful con- trol measures other than precipitation as copper ferrocyanide, which is illegal, have not been developed as yet. Clouding as a result of the formation of colloidal suspensions of in- soluble ferric compounds on oxidation or of insoluble copper compounds on reduction is more likely to occur in dessert wines prepared from must fermented off the skins than from must fermented with the skins and seeds. This may be due either to absorption of tannins or to loss of pro- tective colloids in the latter case. The addition of tannin to wine often but not always stabilizes it against metal casse. Sulfur dioxide in quantities up to 75 parts per million is also bene- ficial owing in part to its antioxidative effect and in part to formation of aldehyde sulfonates. By binding the free aldehyde, sulfur dioxide retards precipitation of tannins, anthocyanins, and other organic constituents which form insoluble complexes with free aldehyde. The binding of free sulfur dioxide by aldehydes and sugars reduces its antioxidative and antiseptic power. If white wines contain excessive amounts of copper, then sulfur dioxide, even in small amounts, exaggerates copper casse. 175 Marsh, G. L. Metals in wine. The Wine Eeview 8(9) :12-13, 24; (10) :24-26, 28-29. 1940. 150 University of California- — Experiment Station Oxidasic Casse. — Oxidasie oasse, the rapid yellowing of white wines and browning of red in the presence of grape oxidase, may occur in dessert as well as table wines. Hussein and Cruess 176 found that grape peroxidase in the presence of 20 per cent alcohol retains over 60 per cent of the activity it has in the absence of all alcohol. Pasteurization, acidi- fication, or addition of sulfur dioxide inhibits grape peroxidase. 177 ANALYSES 178 The chemical analysis of a wine is of great assistance to the wine maker in blending procedures and in deciding upon their rational handling and treatment. It is not at the present time, however, a satisfactory method of deciding upon the quality of a wine although certain types of spoilage may be recognized by their effect on the composition of the wine. The chemical analysis of dessert wines is usually made by chemists who have had considerable experience in wine analysis. The following procedures, which are chiefly adapted for control work, have been selected with this fact in mind. For more detailed discussion of the several methods, consult the references on pages 178 to 179. Some of the procedures outlined require training in analytical chemistry and should not be attempted by those who have not had such training. Other pro- cedures are comparatively simple. ALCOHOL By Ebullioscope. — The method is based on the regular variation in the boiling point of mixtures of alcohol and water with alcohol content. For wines having an extract content (see p. 157) up to 5, accurately dilute the wine 1 : 1 with water, using pipettes to measure the wine and the water. With wines having extract contents from 5 to 10, dilute 2 :1 with water and dilute 3 : 1 for wines having extract contents of over 10. In the presence of sugar, the ebullioscope gives high results since sugar and other nonalcohol soluble solids decrease the boiling point of alcohol- water solutions. 179 The effect of sugar may be reduced by diluting the wine as outlined above, but this introduces other errors. For approxi- mate results the alcoholic content may be determined by suitable dilu- tion. As used by United States gaugers, the wine is diluted to an extract 17,1 Hussein, A. A., and W. V. Cruess. Properties of the oxidizing enzymes of certain vinifera grapes. Food Besearch 5:637-48. 1940. 177 Hussein, A. A., and W. V. Cruess. A note on the enzymic darkening of wine. Fruit Products Journal 19:271. 1940. 178 General references on this subject in addition to those given in specific footnotes in the section are listed on p. 178-79. 170 Love, E. F. A table for ebulliometers. Industrial and Engineering Chemistry, analytical edition 11:548-50. 1939. Bul. 651] Commercial Production of Dessert Wines 151 content of not over 6 per cent and alcohol content of not over 11 per cent. More accurate determinations with the ebullioscope can be made by use of correction tables such as those of Love or of Churchward. 180 Another procedure is to distill the wine and determine the alcoholic content of distillate by ebullioscope. As the boiling point of water varies with the atmospheric pressure, it must be ascertained at the time each series of determinations is made. A sliding scale, on which the boiling point should be adjusted from day to day or even twice a day, accompanies the instrument and is used for calculating the alcohol content. Pipette 15 cc of water into the boiling chamber of the ebullioscope, leave the condenser empty, insert the thermometer, and heat evenly with a carefully adjusted alcohol burner or gas microburner. The tempera- ture reading is taken when the mercury column is constant for about 30 seconds. The temperature reading for this blank determination is set on the sliding scale opposite 0.0 per cent alcohol. The boiling water and condenser water are then drained, the boiling chamber rinsed with a little of the wine to be tested (diluted, if necessary), 50 cc of wine placed in the chamber, the condenser filled with cold water, the ther- mometer replaced, heat applied, and the boiling point determined. The reading is noted on one side of the sliding scale; opposite this tempera- ture the alcohol percentage in the wine (or in the diluted wine) is read. (Churchward prefers to use a specially constructed table for the Dujar- din-Salleron ebullioscope, as is customary for other ebulliometers.) If the wine has been diluted, this is multiplied by the proper factor. This method is usually accurate enough for the alcohol determination on the fermenting wine to be fortified or for routine winery operations. Repli- cate results by the method should check to within ± 0.25 per cent. 151 For more exact results, the following method is recommended. By Distillation. — For analysis by distillation, pipette 100 cc of the wine into an 800-cc Kjeldahl flask and add 50 cc of distilled water. Con- nect to a condenser and distill about 97 cc into a 100-cc volumetric flask. Bring the flask to 68 ° F ( 20 ° C ) and make the volume up to the mark with distilled water. Shake and bring to 68° again. Empty the alcohol into a cylinder and float an alcohol hydrometer in it. Be sure that the hydrom- 180 Churchward, C. E., and B. G. Johns. The use of the Dujardin-Salleron ebulliom- eter for the determination of the alcoholic strength of wines. Australian Chemical Institute Journal and Proceedings 7:18-30. 1940. Churchward, C. E. The Dujardin-Salleron ebulliometer. Australian Chemical Institute Journal and Proceedings 7:71-72. 1940. 181 Joslyn, M. A., G. L. Marsh, and J. Fessler. A comparison of several physical methods for the determination of the alcohol content of wine. Association of Official Agricultural Chemists Journal 20:116-30. 1937. Valaer, P. Eeport on alcohol by the use of the ebullioscope. Association of Official Agricultural Chemists Journal 21:175-77. 1938. 152 University of California — Experiment Station eter is clean and dry and do not touch it except at the top. Carefully make a reading and then take the temperature of the solution. Correct the reading by reference to table 32. (If the hydrometers are calibrated at some other temperature, a different table will be necessary.) Results by this method should check to within ± 0.15 per cent. Chemical Analysis by Acid Dichromate. — Of the several available chemical methods for the determination of ethyl alcohol, the most relia- ble depends on the determination of the quantity of acid dichromate required to oxidize alcohol to acetic acid, 182 or the amount of alkaline permanganate necessary for the complete oxidation of alcohol. 183 The oxidizing mixture is added in measured excess and the quantity remain- ing in solution is determined by back titration with a suitable reducing agent. More recently the excess of oxidizing agent has been determined colorimetrically. 184 For the Semichon-Flanzy procedure, pipette 20 cc of wine of 14 per cent alcohol content or below, or 10 cc of dessert wine into a small 100-cc round-bottom flask and add about 30 cc of distilled water. Distill directly through a small condenser into a 100-cc volumetric flask until a little over half the liquid is distilled over. Make up to volume at 15° C with distilled water. Place 20 cc of potassium bichromate solution (33.832 grams dissolved in water and brought to 1 liter at 15° C) and 10 cc of concentrated H 2 S0 4 (sulfuric acid) in a 250-cc Erlenmeyer flask (preferably a glass- stoppered flask). Mix and then cool in running water to room tem- perature. Add 5 cc of the alcohol distillate to the mixture, mix and stopper the flask with a clean rubber or glass stopper. Let stand at room tempera- ture, with occasional mixing, for 10 minutes. Titrate the excess chromate with the ferrous ammonium sulfate solu- tion (135.310 grams in 700 cc of cold water to which are added 20 cc of concentrated H 2 S0 4 and the solution made to 1 liter; 2 cc of this solu- tion should correspond exactly to 1 cc of the potassium bichromate solution) using a porcelain spot plate with 1 per cent potassium ferri- cyanide as external indicator. The liquid, which is brown at first, soon turns green. As soon as it turns green, add the ferrous ammonium sulfate dropwise and place a drop of the titration mixture on a drop of the 182 Semichon, M. L., and M. Flanzy. Dosage de l'alcool dans les vins et spiritueux par Pemploi du melange sulfi-chromique. Annales des Falsifications et des Fraudes 22:139-52.1929. 183 Friedmann, Theodore E., and Eosalind Klaas. The determination of ethyl al- cohol. Journal of Biological Chemistry 115:47-61. 1936. 184 Gibson, John G., 2d, and Harry Blotner. The determination of ethyl alcohol in blood and urine with the photoelectric colorimeter. Journal of Biological Chemistry 126:551-59. 1938. Bul. 651 Commercial Production of Dessert Wines 153 o o +3 £3 u O a o u o o H fa ^ ■ •* to 5 co co co o o os CONOOOOOS CSOH^OO OCM CO-* to CMt>- CM t^ CO tO CD b~ 00 •* tO CD t^ P. CM CM CM CM CM CM CM COCO CO CO CO CO CO fa oo § ■CCmiOO « CMCOCO^ CD CM OS t-- CO oo a> o T-i co •<*<00 CM CO0O oooo i-H CM CO •>*< fe .i-H^~ P. i-I^CMIMCM CMCN CMCMCM CO CO CO CO fa CO -< § 'OMNH NN00CO'* CM COCO ■* to hOOhW (DWOOCSO CD00 ^H •* to i-l CM •* lO CO CO •*! coco NOOOO CM CM CM CM CM CM CM CM CO | fa < 2 co •* co o co OS CO 00 to CM O-H-HCNCO OS CO to lO CD CO •* to CD t— OO 03 O CM CM O0 OS r+ CO CO i-H 00 CD lO ■* •* tOCO CM CM CM CM P, fa < per cent 0.77 0.78 0.80 0.83 0.86 Oi OS OS O i— < ooo--*^ CD CO >-c OO »-( CM COCO to OOOOO (CNQOOlO ,_;_; ^;^c CO < S *< 00 co co ^ ^h ^ CM ■* OO CM CO lO to to CO CD OOHifN CO tr^ t>. t^ O fa CO CO § CO N OO Oi O O -H CO TfH lO CM CM CM CM CM t^ OS -H CM to CM CM CO CO CO OO-itCCOO) i— l CO to CD tO to to to fc© 0, o fa < § O r-< CM CM CO ■* ■* co co n OOO-lN'* ^H CM CM CM CM (ON050N CM CM CM CO CO ■* cot^oo CO CO CO CO o fa CO < S to to co co n Sooooo NN O0 OO OO ooooo OiOOrtN CO-* •* to CD t^ t^ oooo p. o fa to § t^ to to to to Rococo ■to co co e^. oo ©o oooo So SO -3— s- o fa to < u >-< 1-1 •—< >-l 1-H ?-- 00 Oi »-, so 1-1 »-1 "I ©i ©* ©} «X ©J so so *^0 >*^*H 60^*-^--*tO So Co 00 o >o to to CO 1° o lOOfl ^ ■.'.;■. ; ; : ; ; : : : : ; ; : : : : : : : : : rtNMM'W (ONOOOO — I CM CO'* tO CON 0000 -H CM CO 1 * CM - Jo as .9 'it a m CD 154 University of California — Experiment Station ferrocyanide; as long as there is an excess of chromate, this latter solu- tion will turn a yellowish orange in color; as soon as the ferrous salt is in excess, the indicator turns to a characteristic blue. The solution in the flask will become blue-green if it has an excess of ferrous solution. The alcohol content in the wine is calculated as follows : If n is the number of cubic centimeters of the ferrous ammonium TABLE 33 Specific Gravity of Alcohol Solutions at 60° F (Assuming specific gravity of water at 60° F as unity) Alcohol Specific gravity Alcohol Specific gravity per cent 0.0 1.00000 0.99925 .99850 .99776 .99703 .99630 .99559 .99488 .99419 .99350 .99282 .99215 .99150 .99085 .99022 .98960 .98899 .98838 .98779 .98720 .98661 .98602 0.98544 per cent 11 5 12 98487 5 98430 1 12 5 98374 15 13 98319 20... . 13 5 98264 2 5 ... 14.0 14 5 98210 3.0 . 98157 3.5. . 15 0. . 98104 4 . 15 5. . 98051 4 5 . 16.0 16.5 17 0. . . 97998 5.0 .97946 5 5 . 97895 6 17 5 97844 6 5 . 18 0.. 97794 7 . 18 5 97744 7.5 19.0 .97694 8.0 19.5 .97645 8.5 20.0 .97596 9.0 20.5 .97546 9.5 21.0 .97496 10.0 21.5 22.0 .97446 10.5 .97395 11.0 22.5 0.97344 Source of data: United States Bureau of Internal Revenue. Regulations No. 7 relative to the production, fortification, tax payment, etc., of wine. 188 p. United States Government Printing Office, Washington, D. C. 1937. sulfate solution added to reduce excess chromate, then the alcohol con- tent of the wine itself is (20 — ) for the 20 cc of dry wine used, or double Li that for the 10 cc of sweet wine used. For example if n = 8.50, then alcohol content, in per cent by vol- .50 ume = 20 - 20-4.25, or 15.75. The method depends on the complete oxidation of alcohol to acetic acid. The acid, alcohol, and chromate concentrations, and time at room temperature must be carefully adhered to in order to achieve complete oxidation of alcohol to acetic acid without oxidizing the latter further. Bul. 651] Commercial Production op Dessert Wines 155 The concentrations of reagents nsed have been so adjusted that at 15° C, 1 cc of dichromate solution completely oxidizes 7.943 mg of alcohol to acetic acid and corresponds to 1 per cent by volume of alcohol in the orig- inal wine. The presence of reducing matter other than alcohol in distillate obviously interferes with the determination. Usually this is small for French wines, and with the average California wine the recovery of alcohol by this method is accurate to ± 0.05 per cent alcohol. The dichromate solution is fairly stable but the ferrous ammonium sulfate solution will deteriorate unless protected from oxidation as by charging with ordinary gas or hydrogen. Modified Dichromate Procedure. — The above procedure may be modi- fied by the use of an internal indicator such as diphenylamine, or di- phenylamine sulfonic acid 185 instead of the external indicator. Proceed as above until the titration. To the mixture of excess dichromate and acetic acid or to an accurately measured aliquot, add a measured excess (25 to 50 cc) of ferrous ammonium sulfate solution. (It is better to make up the mixture to 100 cc and pipette out 25 cc, to which is added the acid ferrous ammonium solution.) Then add 4 drops of diphenylamine indi- cator (1 gram of diphenylamine dissolved in 100 cc of concentrated H 2 S0 4 ) and titrate the excess of ferrous salt with dichromate solution. In the presence of excess ferrous salt, the solution is green; as standard dichromate solution is run in, the green color changes to a blue-green and then to an intense blue or violet color. Titrate the volume of ferrous ammonium sulfate solution added in the above step with the dichromate solution. Add the same number of drops of indicator and 5 cc of con- centrated sulfuric acid to the measured volume of ferrous ammonium sulfate solution. The diphenylamine end point is improved by the addition of 15 cc of phosphoric acid mixture (150 cc of concentrated H 2 S0 4 + 150 cc HoP0 4 sirup diluted to 1 liter) to bind the ferric ion complexes : 1S6 Calculate the weight in grams of the alcohol present in the aliquot used for titration from the reaction : 2K 2 Cr 2 7 + 3C 2 H 5 OH + 8H 2 S0 4 == 2Cr 2 (S0 4 ) 3 + 3CH 3 C0 2 H + 2K 2 S0 4 + 11 H 2 0. Then the concentration of alcohol present in the distillate is calculated 185 Sarner, L. A., and I. M. Kolthoff. Diphenylamine sulfonic acid as a new oxi- dation-reduction indicator. American Chemical Society Journal 53:2902-05. 1931. Sarner, L. A., and I. M. Kolthoff. Indicator corrections for diphenylamine, diph- enylbenzidine, and diphenylamine sulfonic acid. American Chemical Society Journal 53:2906-9.1931. 186 Bottger, William (Translated by Ralph E. Oesper). Newer methods of volumet- ric chemical analysis. 268 p. (See particularly section on oxidation-reduction indica- tors by Erna Brennecke. p. 155-80.) D. Van Nostrand Company, Inc., New York, N. Y. 1938. 156 University of California — Experiment Station as grams per 100 cc, and converted into percentage of alcohol by volume by reference to table 17 in Bulletin 652. Another modification of the dichromate procedure has been developed by Fessler. 187 The specific gravity of alcohol solutions from 0.0 to 22.5 per cent is given in table 33. ■»«■.»• . « TO fe TOTAL ACID For white wines, pipette 10 cc of wine into a 500-cc flask, and if carbon dioxide is present, add 250 cc of boiling distilled water. Add 3 to 5 drops of phenolphthalein indicator solution; and titrate with standardized 0.1 N NaOH (0.1 normal. sodium hydroxide) to a distinct pink color, or until 1 or 2 drops of NaOH produce no perceptible change. One may best observe the color change by holding the flask just above a well- lighted white surface. The total acidity, expressed as grams of tartaric acid per 100 cc, is obtained by multiplying the number of cubic centi- meters of 0.1 N NaOH used in titrating by 0.075. For dark and red wines, measure 2.0 cc of wine into a 500-cc flask. Add, if carbon dioxide is present, 250 cc of boiling distilled water and several drops of phenolphthalein indicator. Disregard the red to bluish- black color change, and titrate with 0.1 N NaOH to a pink end point. The total acidity expressed as grams of tartaric acid per 100 cc is obtained by multiplying this titration figure in cubic centimeters by 0.375. For this titration, a microburette is preferred, although a 10-cc burette graduated in 1/20-cc intervals is satisfactory. More dilute NaOH may be used also, for example, 0.033 N. In this case a regular burette may be used. VOLATILE ACID The method of analysis for volatile acid is based on the fact that certain of the acids formed in large amounts in spoiled wine (mainly acetic) are distilled by steam at atmospheric pressure. With a 10-cc pipette, introduce 10 cc of wine into the central tube of a volatile-acid distillation apparatus. Add 150 cc of recently boiled hot distilled water to the outer flask. Tightly connect the apparatus as directed by the manufacturer. Start cold water running through the condenser, and apply heat to the outer flask. When the water has boiled for a moment, close the pinchcock on the outlet of the outer flask, and distill until 100 cc has been collected in a 300-cc flask. Heat the distillate to boiling, add 3 to 5 drops of phenolphthalein solution, and titrate with 0.1 N NaOH. The volatile acid content as grams of acetic acid per 100 cc of wine is obtained by multiplying the titration in cubic centimeters by 0.06. Somewhat more accurate results may be obtained for the titration by using a more dilute NaOH solution. 187 Fessler, J. H. Alcohol determination by dichromate method. Wines and Vines 22(4): 17-18. 1941. Bul. 651 Commercial Production of Dessert Wines 157 EXTRACT Pipette 100 cc of wine into a 250-cc beaker, add 50 cc of distilled water, place over a moderate source of heat, and evaporate to approximately 50-cc volume. Remove from the heat and cool to room temperature. Dilute back to the original volume in a 100-cc volumetric flask by adding distilled water and mix thoroughly. Adjust the temperature to exactly 20° C and then place a portion of the liquid in a glass hydrometer cylin- TABLE 34 Corrections for Brix or Balling Hydrometers Calibrated at 20° C (68° F) Temp of so erature Observed percentage of sugar ution 5 10 15 20 25 30 Below calibration Subtract ° C 15 ° F per 59.0 cent 20 per cent 0.22 per cent 24 per cent 0.26 per cent 0.28 per cent per cent 30 0.32 15.56 60.0 18 .20 .22 .24 .26 .28 .29 16 60.8 17 .18 .20 .22 .23 .25 .26 17 62.6 13 .14 .15 .16 .18 .19 .20 18 64.4 09 .10 .11 .12 .13 .13 .14 19 66.2 05 05 0.06 0.06 0.06 0.07 0.07 Above calibration Add °C 21 ° F per 69.8 cent 04 per cent 05 per cent 06 per cent 0.06 per cent 07 per cent 07 per cent 0.07 22 71.6 10 .10 .11 .12 .13 .14 .14 23 73.4 16 .16 .17 .17 .20 .21 .21 24 75.2 21 .22 .23 .24 .27 .28 .29 25 77.0 27 .28 .30 .31 .34 .35 .36 26 78.8 33 .34 .36 .37 .40 .42 .44 27 80.6 40 .41 .42 .44 .48 .52 .52 28 82.4 46 47 .49 .51 .56 .58 .60 29 84.2 54 .55 .56 .59 .63 .66 .68 30 86.0 61 0.62 0.63 0.66 0.71 0.73 0.76 35 95 99 1.01 1.02 1.06 1.13 1.16 1.18 Source of data: Association of Official Agricultural Chemists. Official and tentative methods of analysis. 5th ed. p. 664. Washington, D. C. 1940. der. Read off the approximate extract with a Brix or Balling hydrometer. When the dealcoholized wine has been adjusted to the same volume as that of the original sample, the reading gives the approximate extract direct as grams per 100 cc of wine. The hydrometer reading will have to be corrected for the influence of temperature, if the reading is made at a temperature different from that at which the hydrometer is calibrated. (See table 34.) If specific-gravity hydrometers are used, the readings may be converted to degree Balling by reference to table 35. 158 University of California — Experiment Station If the alcohol has been determined by distillation, the solution remain- ing in the Kjeldahl flask may be used for the extract determination. The solution is poured into a 100-cc volumetric flask. The sides of the Kjel- dahl flask are washed down and this solution also poured into the volu- metric flask. This is repeated several times until the extract is all removed TABLE 35 Specific Gravity at 60° F Corresponding to Eeadings of the Balling Hydrometer (Assuming specific gravity of water at 60° F as unity) Balling Specific gravity Balling Specific gravity Balling S g oecific •avity degrees 0.00 1.0000 degrees 10.0 1 03933 degrees 20.0 1 0814 0.50 1.0019 10.5 1 0414 20.5 1 0836 1.00 1.0038 11 1 0434 21 1 0859 1.50 1 00575 11 5 1 0454 21.5 1 0881 2.00 1 0077 12.0 1 04746 22 1 0903 2.50 1.00966 12.5 1 0495 22.5 1 0926 3 00 1.01163 13 1.05153 23.0 1 0949 3 50 1 01356 13.5 1.05356 23.5 1 0971 4 00 1.0155 14 .0 1.05556 24 1 0994 4 50 1.0174 14 5 1.05753 24 5 1 1017 5.00 1.01926 15.0 1 05943 25.0 1 1040 5 50 1.0212 15 5 1.06163 25.5 1 1063 6.00 1.02323 16.0 1.06386 26.0 1 1086 6.50 1.0252 16 5 1 0660 26.5 1 1109 7.00 1.02706 17.0 1.06803 27.0 1 1133 7 50 1.02906 17 5 1 0701 27.5 1 1155 8.00 1.0313 18 1.07233 28.0 1 1180 8 50 1 03336 18 5 1.07456 28.5 1 1203 9.00 1.03523 19 1.0769 29.0 1 1227 9 50 1 0372 19 5 1 07926 29.5 1 30.0 1 1251 1274 Source of data: United States Bureau of Internal Revenue. Regulations No. 7 relative to the production, fortification, and tax payment, etc., of wine. 188 p. United States Government Printing Office, Washington, D. C. 1937. from the Kjeldahl flask. The volumetric flask is then brought to 20° C, made to volume with distilled water, and shaken, and the extract content determined with the hydrometer as indicated above. The extract content may also be obtained by pipetting 50 or 100 cc of the wine into a tared platinum or porcelain dish, evaporating off the moisture at 70° C in a vacuum oven, and weighing. REDUCING SUGARS The extract content minus 2.5 usually gives the sugar content accu- rately enough in the sweet dessert wines. Occasionally in sherries it is desirable to know the actual reducing-sugar content. The volumetric method given below is based on the Lane and Eynon titration method for determining substances capable of reducing copper Bul. 651] Commercial Production of Dessert Wines 159 in alkaline tartrate solution. Although the copper-reducing substances in wine are largely sugars, other reducing matter is present and must be removed. The standard procedure is to clarify with lead acetate and then remove the excess lead with sodium oxalate. The treatment with charcoal given here is simpler though less accurate. White wines contain much less of interfering substances than red wines. Dealcoholize the wine (as for extract), cool, and make to the original volume by adding distilled water. With white wine, filter if not bril- liantly clear. With red wine, decolorize by shaking 100 cc of wine with about 5 grams of acid-washed decolorizing carbon and then filtering. Immediately before use, prepare the Soxhlet reagent by mixing 50 cc each of Fehling's A and B solutions 188 in a clean, dry flask. Pipette accurately 25 cc of the mixed Soxhlet reagent into a clean 300-cc Erlenmeyer flask. To standardize the method, fill the burette with a standardized 0.5 per cent dextrose solution and also pipette 20 cc of this solution into the flask. Set the flask on a wire gauze and place the burette just above the mouth of the flask. Heat the cold mixture to boiling and maintain a moderate boiling for about 15 seconds; lower the flame enough to avoid bumping. Add rapidly further quantities of the sugar solution from the burette until only the faintest perceptible blue color remains. Without removing the flame, add 2 to 5 drops of 1 per cent methylene blue solution ; and titrate dropwise until the indicator is com- pletely decolorized. The total volume of the standard solution necessary should be about 24 cc. To determine the sugar content of the wine, proceed as above, adding 20 cc of the prepared wine solution to 25 cc of the mixed Soxhlet reagent. After boiling it for 15 seconds, finish the titration by adding standard dextrose solution from the burette. If the 20 cc of wine solution com- pletely decolorizes the Soxhlet reagent, dilute the wine solution further, and repeat the determination. For example, take 20 cc, dilute to 100 cc, and take 20 cc of the diluted liquid, multiplying the result by 5. From the amount of the standard dextrose solution necessary to reduce the copper completely and from the amount necessary to finish the reduction after the addition of the wine solution, the sugar content of the latter can be calculated as follows : per cent of dextrose = Qn n , where a is the cubic centimeters of dextrose required for direct titration and b is the cubic centimeters of dextrose required for the back titration. 188 Fehling's A is made by dissolving 34.639 grams of copper sulfate (CuS0 4 5 H 2 0) in water and making to 500 cc. Fehling's B is made by dissolving 173 grams of sodium-potassium tartrate (Rochelle salts), and 50 grams of NaOH in water and making to 500 cc. 160 University of California — Experiment Station TANNIN AND COLORING MATTER 188 The method of analyzing for tannin and coloring matter depends on the determination of the permanganate reducing matter of the wine before and after decolorizing with carbon. Dealcoholize the wine, cool, and make to the original volume. Trans- fer 5 cc to an 800-cc beaker. Add about 500 cc of water and exactly 5 cc of indigo solution (made by dissolving 6 grams of indigo — free of indigo blue — in 500 cc of water and 50 cc of H 2 S0 4 and making to 1 liter). Titrate with 0.1 N KMn0 4 (potassium permanganate) solution, 1 cc at a time, until the blue color changes to green ; then add a few drops at a time until the color becomes golden yellow. Thoroughly stir the solution with a glass rod or an electric stirrer during the titration. Designate the cubic centimeters of KMn0 4 solution required to reach the golden- yellow color, using the dealcoholized wine, as a. Decolorize and detannize a portion of the dealcoholized sample by shaking well with carbon. (This charcoal should be free of reducing substances.) Filter and titrate 5 cc, by the same procedure previously used, with KMn0 4 . Designate the number of cubic centimeters of KMn0 4 used for the dealcoholized, decolorized, and detannized sample as ~b. This volume is fairly constant and for routine work need be determined only in an occasional sample. Then c, the number of cubic centimeters of KMn0 4 solution required for oxidizing the tannin and coloring matter in 5 cc of the wine, may be calculated from c = a-b. The amount of tannin as grams per 100 cc of wine is equal to c X normality of KMn0 4 X 0.0416 X — ; — 10 °,. . . volume oi wine Recently a more specific colorimetric method for determining the tannin content of whiskeys has been proposed. 100 This may also be suit- able, after modification, for wine. ALDEHYDE The aldehyde content of wine may be determined by the following direct iodometric procedure, although for more accurate results (partic- ularly in highly sulfited wines) the indirect procedure of Jaulmes and Espezel 1 " 1 as given in Bulletin 652 for brandy should be used. For free aldehyde distill 100 cc of wine with 50 cc of water into a 300-cc Erlen- 180 See also p. 20. 190 Eosenblatt, M., and J. V. Peluso. Determination of tannins by photocolorimeter. Association of Official Agricultural Chemists Journal 24:170-81. 1941. 101 Jaulmes, P., and P. Espezel. Le dosage de l'acetaldehyde dans les vins et les spiritueux. Annales des Falsifications et des Fraudes 28:325-35. 1935. Bul. 651] Commercial Production of Dessert Wines 161 meyer flask placed in an ice bath until about 100 cc of distillate has been obtained. For total aydehydes add 5 cc of 85 per cent phosphoric acid to the wine in the distilling flask before distillation. Joslyn and Comar give the following method for determining the aldehyde content in the distillate : Mix 100 cc. of the aldehyde solution, 10 cc. of 0.1 N sodium bisulfite solution con- taining 10 per cent of ethyl alcohol by volume, and 10 cc. of alcohol (if the sample contains none) in a 300-cc. Erlenmeyer flask which is stoppered and allowed to stand at room temperature for 30 minutes. (Aldehyde solutions obtained by distillation were cold when bisulfite was added, so that the solutions were not at room temperature during the whole of the storage period.) Then add 10 cc. of 0.1 N iodine solution from a freely flowing pipet, and back-titrate the excess of iodine with 0.1 N thio- sulfate solution. As a blank, to the same volume of water and alcohol add 10 cc. of the bisulfite solution, let stand for 30 minutes, add iodine, and back-titrate as above (1 cc. of 0.1 N thiosulfate is equivalent to 0.0022 gram of acetaldehyde). 192 ESTERS The esterification of wine has been studied in some detail by Peynaud and others. 193 Peynaud 104 points out that in the usual distillation proce- dure acetaldehyde introduces an appreciable error in the determination of esters. To reduce this source of error, Espil and Peynaud 195 suggest that an aliquot (50 to 100 cc) of the wine be neutralized with 0.1 iVNaOH to exactly pH 7 and then extracted in a liquid-liquid continuous extractor with petrol ether, 25 cc of 0.1 N NaOH being added to the receiving flask so that saponification and extraction occur continuously. The ester con- tent is determined from the back titration of the alkali remaining. The extraction usually requires 12 hours. For more quickly determining the ester content, see the procedure for brandies in Bulletin 652. COLOR The standardization of color is commonly made by direct visual com- parison of the samples. This is satisfactory if an original sample is available, but in most cases it is necessary to specify the color in terms of some permanent physical or chemical standard. The only completely satisfactory method of specifying color is to measure the transmission 192 Joslyn, M. A., and C. L. Comar. Determination of acetaldehyde in wines. Indus- trial and Engineering Chemistry, analytical edition 10:364-66. 1938. 193 Eibereau-Gayon, J., and E. Peynaud. Cited in footnote 39, p. 20. Peynaud, E. Les phenomenes d'esterification dans les vins. Annales des Fermenta- tions 3:242-52. 1937. Espil, L., L. Genevois, E. Peynaud, and J. Eibereau-Gayon. Sur la formation des esters de l'alcool ethylique. Enzymologia 4:88-93. 1937. 104 p e y nau( ^ e. Etudes sur les phenomenes d'esterification dans les vins. Eevue de Viticulture 86:209-14, 227-30, 248-53, 299-301, 394-95, 420-23, 440-44, 472-73; 87:49-51, 113-16, 185-87, 242-49, 278-85, 297-301, 344-50, 362-64, 383-85. 1937. 195 Espil, L., and E. Peynaud. Dosage des esters neutres dans les milieux de fer- mentation. Societe Chimique de Prance Bul. 3(serie 5) : 2324-25. 1936. 162 University of California — Experiment Station of light through the wine at various wave lengths of the visible spectrum. This method requires costly equipment and is time-consuming. 1 " All the basic color characteristics, however, are specified : dominant wave length, brilliance, and purity. The Dujardin-Salleron vino-colorimeter is useful only for red wines and only young ports may be tested, as the color disks for specifying dominant wave length are not of the same hue as aged ports. For young ports, however, a rough measure of the domi- nant wave length and the depth of color (brilliance) may be obtained by the use of this instrument. The color dyes recommended by Winkler and Amerine are likewise not well adapted to aged ports unless the pro- portions are changed to include more brown so that a better color match may be obtained. A fairly permanent color standard for white wines, particularly for sherries, may be made from the Eastman ABC dyes. Ten cc of red and 5 cc each of blue and yellow from 0.5 per cent solutions when made up to 300 cc with water provides a convenient empirical standard. The use of colored slides of the Lovibond apparatus for specifying color is a further possibility. Both red and white wines may be tested. Although the color data obtained represent arbitrary units, they may be compared with each other and therefore do provide a permanent record of the color characteristics of the wine. They should therefore prove useful in blending tests. Photoelectric colorimeters or photoelectric spectrophotometers are particularly useful in measuring and recording color in wines. The transmission should be measured with several color filters; the filter showing greatest variation should be used for primary blending. IRON DETERMINATION 197 Total iron in wines is most easily and accurately determined by the procedure developed by Saywell and Cunningham. 198 The method in- volves wet-ashing and the development of a colored complex which fer- rous iron forms with o-phenanthroline. Pipette 2 cc of wine into 25 X 150 mm Pyrex test tubes previously marked at 10 cc. Evaporate to dryness, cool, and add 1 cc of concentrated H 2 S0 4 . Heat over a flame under a hood with care until the contents of the loo Winkler, A. J., and M. A. Amerine. Color in California wines. I. Methods for measuring color. Food Eesearch 3(4) :429-38. 1938. 197 This section was prepared by George L. Marsh, Associate in the Experiment Station. 108 Saywell, L. G., and B. B. Cunningham. Determination of iron. Colorimetric o-phenanthroline method. Industrial and Engineering Chemistry, analytical edition 9:67-69. 1937. See also : Hummel, F. C, and H. H. Willard. Determination of iron in biological materials. The use of o-phenanthroline. Industrial and Engineering Chemistry, ana- lytical edition 10:13. 1938. Bul. 651] Commercial Production of Dessert Wines 163 tube are completely liquefied. Allow to cool, and then add 0.5 cc of 70 per cent HC10 4 (perchloric acid). Heat continuously until partial clari- fication has occurred, set aside to cool, and then add an additional 0.5 cc of HC10 4 . Continue the digestion until the sample is clear and until all the excess HClO± has been evaporated off. At this stage set the tubes aside to cool. Caution: the digestion should be conducted behind a shat- terproof glass, because perchloric acid occasionally explodes during heating. Add 2 cc of distilled water and a small piece (0.5 cm square) of Congo red paper. Then add 1 cc of a 10 per cent aqueous solution of hydroxyla- mine hydrochloride and 1 cc of a 0.1 per cent solution of o-phenanthro- line in 50 per cent alcohol. Titrate to the color change (blue to a light red) of the Congo red paper with concentrated NH 4 OH and set aside to cool. Then make to 10 cc with distilled water and transfer to standard- ized test tubes for comparison against a series of standards or to colorim- eter tubes for comparison by photoelectric photometers. Prepare a standard stock solution of iron to contain 1 mgm iron per cc. 199 Prepare from this standard stock solution, a series of solutions containing known concentrations (in parts per million of iron). Run these solutions through the procedure given above for the unknown, using all the reagents and following directions closely. Transfer to standard-diameter test tubes and cork tightly for use as a series of standards. These standards are stable for long periods and the procedure outlined automatically corrects for the iron in the reagents. In like manner, the solutions can be used to establish a curve for use with a photoelectric colorimeter. In this case, however, a blank containing water instead of an iron solution but containing all the reagents must be prepared. Use this blank to set the instrument to zero reading. COPPER DETERMINATION 200 Quantitative. — Copper in wine is most accurately determined by a modified Coulson 201 -Drabkin 202 procedure. It is a colorimetric test using sodium diethyldithiocarbamate as the reagent for color production. When used under the conditions of the test, the reagent is specific for copper and forms a reasonably stable color complex which can be com- pared against a series of standards by visual examination or in colorim- 199 Weigh out 7.022 grams ferrous ammonium sulfate, dissolve in 500 cc of distilled water to which 5 cc of concentrated HC1 has been added. Transfer to a 1,000-cc volumetric flask and make to volume with distilled water. 2,10 This section was prepared by George L. Marsh, Associate in the Experiment Station. 301 Coulson, E. J. Report on copper. Association of Official Agricultural Chemists Journal 20:178-88. 1937. 202 Drabkin, D. L. Report on copper. Association of Official Agricultural Chemists Journal 22:320-33. 1939. 164 University of California — Experiment Station eters of the Duboscq type, or against standard curves in photoelectric photometers. Pipette 25 cc of wine into 100-cc Kjeldahl flasks. Place the flasks in a hot air oven at 100° C and heat until the sample is dry. Remove from the oven and when cool add 5 cc of concentrated H 2 S0 4 and 5 cc concen- trated HN0 3 . Heat gently until red N0 2 fumes appear and set aside overnight. Then heat gently at first and more strongly as frothing ceases to the complete disappearance of N0 2 fumes or to the appearance of S0 3 fumes. Set aside to cool and when cool add 1 cc of 70 per cent HC10 4 . Again heat until the fumes disappear and there is complete clearing of the solution. Cool. Eapidly add about 5 cc of distilled water and transfer the contents of the Kjeldahl flask to a 150-cc beaker. Wash out the flask and transfer the washings to a beaker, using the minimum amount of water necessary. Add 10 cc of hydrochloric-citric acid reagent. 203 Then neutralize with concen- trated NH 4 OH to litmus and add 0.2 cc excess. Add 1 cc of a 1 per cent aqueous solution of sodium diethyldithiocarbamate and transfer the con- tents of the beaker by careful washing to a 125-cc pear-shaped separatory funnel which has been previously marked at 50 cc. Make up to this volume with distilled water. Then accurately pipette 10 cc of isoamyl acetate over the solution, stopper, and shake with reasonable vigor for 1 minute. Allow the two liquid phases to separate, draw off the aqueous phase, dry the stem of the separatory funnel with a pipe cleaner, and then transfer the organic phase of the solution to the colorimetric comparator tubes. Standard copper solution is best prepared from reagent copper metal (reagent quality) so as to contain 1.25 mg copper per cc. Solution for the preparation of the series of standards is prepared by pipetting 5 cc of the above solution into a 500-cc volumetric flask and making up to volume with distilled water. Pipette 1, 2, 3, 4, 5, and 6-cc aliquots of the latter solution into 100-cc Kjeldahl flasks. Add 5 cc of concentrated H 2 S0 4 and 5 cc of concentrated HNO3 and boil until all the HN0 3 has been dispelled. Cool, add 1 cc HCIO4 and again boil to dispel this latter acid. Again allow to cool and then quickly add 5 cc of distilled water. Transfer the contents of the Kjeldahl flask to 150-cc beakers and proceed as directed above for the unknown. Draw off the extracted colored solution into tubes of the same size and diameter as those used for the unknown and tightly stopper with corks. Compare the unknowns against the standards in a Walpole block or other convenient comparator. Standards prepared as above 203 For hydrochloric-citric acid reagent, dissolve 75 grams C. P. citric acid in 350 cc of distilled water. Add 50 cc of concentrated HC1, stir, and transfer to a 500-cc volumetric flask and make to volume. Bul. 651] Commercial Production of Dessert Wines 165 described automatically correct for copper in the reagents and are stable for at least 1 month. These standards contain 0.0125, 0.025, 0.0375, 0.050, 0.0625, and 0.075 mg of copper per 10 cc of colored solution, which under the conditions of the test corresponds to 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 parts per million of copper in the wine, respectively. If the unknown contains more than 3.0 parts per million of copper, repeat the determination using a smaller aliquot as sample, but make due allowance for the new volume of the aliquot when calculating the copper concentration. Where a Duboscq colorimeter is available, only two standards at the most need be prepared, one using 1.5 cc of the diluted standard solution, the other using 3.0 cc. At this concentration of copper per cc of colored solution covered by the range of standards above, the color intensity is proportional to the concentration and balancing methods yield reliable values. A photoelectric colorimeter may also be used. Marsh Procedure. — The Coulson-Drabkin procedure is time-consum- ing, and careful attention to details is required for accurate results. Nearly as accurate values can be obtained by the following method in a very much shorter period of time and this one is better suited to the average winery laboratory. Wet-ashing is avoided and other innovations are added for purposes of simplification. Pipette 10 cc of wine into a 25 X 150-mm Pyrex test tube. Add 1 cc of hydrochloric-citric acid reagent, shake, and then add 2 cc of 5iV NH 4 OH (333 cc concentrated NH 4 OH per liter), and again shake. Then add 1 cc of a 1 per cent aqueous solution of sodium diethyldithiocar- bamate, shake and set aside for a minute or so before adding 10 cc of amyl acetate. Follow the addition of amyl acetate with 5 cc of absolute methyl alcohol. Place the palm of the hand over the top of the test tube and shake vigorously for at least 30 seconds. Set aside and allow the two phases to separate. When separation is complete, draw off the aqueous phase by inserting a length of glass tubing 204 to the bottom of the test tube and apply suction. Then dry the organic phase by adding anhydrous sodium sulfate, powdered, from the tip of a spatula and shaking, adding only a sufficient amount to accomplish the purpose. It should be added while holding the tube at an angle and at the tame time rotating the tube for the purpose of drying the moisture film adhering to the walls. Transfer the dried organic phase to clean dry test tubes for color comparison against a set of standards or to the colorimeter tubes of the Duboscq or any of the new photoelectric colorimeters. 204 A soda-glass stopcock, drawn to a capillary on one side and connected to a liter filter flask through rubber tubing on the other side, serves the purpose excellently. Connect the filter flask to a water aspirator or vacuum pump. 166 University of California — Experiment Station The absolute methyl alcohol which is added to the reaction mixture serves two purposes. It markedly reduces the tendency of the two phases to emulsify and thereby aids in a quick and clean separation of the aqueous and organic phases. More important, however, is its second pur- pose. Coulson shows that the intensity of the color which is extractable by amyl acetate from the aqueous phase is dependent upon the pH value of the aqueous phase. At pH 8.0 to 8.5, maximum color intensity is devel- oped and great care must be exercised in adjusting the pH value if ac- curate values are to be obtained. This is common to most colorimetric procedures and Coulson's method differs little from most others in this respect. It was found in developing the procedure outlined above that when methyl alcohol was added to tubes containing samples adjusted to varying pH values, no difference in color intensity of the extracted colored solution could be detected. Without methyl alcohol, the color intensity of the extracted colored solution was dependent upon the pH value of the aqueous phase. With methyl alcohol present the pH value of the aqueous phase can vary over rather wide limits with no effect on the color intensity. Standard copper solution for this procedure is best prepared from Merck copper metal (reagent quality) so as to contain 0.50 mg of copper per cc. Solutions for the preparation of the series of standards are pre- pared by pipetting 1, 2, 3, 4, 5, 6, 8, and 10 cc of the above stock solution into separate 1,000-cc volumetric flasks, adding 150 cc of 95 per cent ethyl alcohol and making each up to volume with distilled water. These solutions then contain 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, and 5.0 parts per million copper, respectively. Pipette 10 cc of each solution into separate 25 X 150-mm test tubes and proceed as directed for the unknown. The solutions so obtained can be used as a series of standards or can be used to establish standard curves for use with photoelectric photometers. SPECIAL TESTS Heating wines containing levulose causes the formation of hydroxy- methylfurfural. This substances gives color reactions with several sub- stances, including rescorcinol. To test for it, shake 10 cc of wine with 10 cc of absolute ether and decant the ether carefully into a small porce- lain evaporating dish. Repeat. Evaporate the ether off, and add 2 cc of freshly prepared 1 per cent rescorcinol (in concentrated HC1), and add 5 cc of concentrated HC1. If hydroxymethylfurfural is present, the contents of the evaporation dish turn red. The amount and depth of color is a rough measure of the hydroxymethylfurfural present. A test can also be made directly in wine with phloroglucinol. In the Bul. 651] Commercial Production of Dessert Wines 167 case of white wines, the original wine may be used, but red wines must be decolorized with bone charcoal (not with activated charcoal). Place 2 cc of the white or decolorized wine in a test tube, add 1 cc of a freshly prepared 1 per cent phloroglucinol solution (made in 30 per cent HC1) and 2 cc of 30 per cent HC1, and shake. Those wines developing a red or deep-orange color contain hydroxymethylfurfural. A method based on the insolubility of the phloroglucinol precipitate has been developed by Kruisheer and co-workers 205 for the quantitative determination of hydroxymethylfurfural. The procedure of Kniphorst and Kruisheer (cited in footnote 18, p. 15) may be used to determine 2, 3 butylene glycol. Glycerin may be determined by the procedure of Bertram and Rut- gers, 206 particularly after separation from the sugar by treatment with lime and extraction with alcohol-ether solution, as indicated in the Of- ficial Method of the Association of Official Agricultural Chemists. Sulfur dioxide may be determined by the procedure given in Bulletin 639, page 108. INTERPRETATION OF RESULTS Certain analytical determinations in dessert wines are of importance from the standpoint of taxation and should be made with care. Usually the very exact determination of all of the constituents of a wine is un- necessary and time-consuming. Small differences in the amounts of the major constituents of wines have but negligible effects on the quality of the wine. Unless required for taxation purposes as mentioned above, results which can be duplicated within the ranges shown below should be satisfactory. Permissible limits of error Substance in determinations Volatile acid ± 0.006 gram per 100 cc Total acid ± 0.03 gram per 100 cc Alcohol ± 0.25 per cent (should be less for procedures other than the ebullioscope) Extract ±0.2 per cent Acetaldehyde ±1.0 mg per 100 cc Balling ± 0.2° Tannin ± 0.01 gram per 100 cc Esters ±3.5 mg per 100 cc Sulfur dioxide, total ±1.0 mg per 100 cc Sulfur dioxide, free ±0.5 mg per 100 cc 205 Kruisheer, C. L, N. J. N. Vorstman, and L. C. E. Kniphorst. Bestimmung des Oxymethylfurfurols und des Lavulosins in Portwein und anderen Sussweinen. Zeit- schrift fur Untersuchungen der Lebensmittel 69 : 570-82. 1935. ^Bertram, S. H., and R. Rutgers. The estimation of glycerol and some other hydroxylated compounds. Recueil Travaux Chimiques des Pays-Bas 57:681-87. 1938. 168 University of California — Experiment Station The analyst need only determine if he is using a standard method which will give results comparable with those commonly reported in the litera- ture, and if his duplicate results fall within the ranges suggested. If his results do not agree with those obtained by standard methods or if duplicate analyses fail to agree, he should change his method or tech- nique or equipment. ACKNOWLEDGMENT We wish to thank the officials of the Wine Insti- tute for their generous cooperation and assistance in the publication of this manuscript. Bul. 651] Commercial Production of Dessert Wines 169 SELECTED REFERENCES FOR FURTHER READING 207 References on the Economic Status of the Industry Blair, E. E., and H. C. Phillips. 1940. Acreage estimates California fruit and nut crops as of 1939. 29 p. California State Printing Office, Sacramento, Calif. Distilled Spirits Institute. 1939. Public revenues from alcoholic beverages. 55 p. Distilled Spirits Institute, Inc., Washington, D. C. Douarche, L., and C. Penic. 1939. reexportation des vins de France. 161 p. Editions de la Journee Vinicole, Montpellier, France. Klatt, Werner. 1932. Die Verwertung der Deutschen Rebenernten. 170 p. Institut fur Landwirt- schaftliche Markforschung, Berlin, Germany. Meigs, Peveril, 3d. 1941. Current trends in California orchards and vineyards. Economic Geography 17(3):275-86. Shear, S. W. 1935. Factors determining wine consumption in the United States. Proceedings of the 3rd Wine Conference. Wine Institute, San Francisco, Calif. (Mimeo.) Shear, S. W., and Gerald G. Pearce. 1934. Supply and price trends in the California wine grape industry. Pt. 2. A sta- tistical survey. University of California Giannini Foundation of Agricul- tural Economics Mimeographed Rept. No. 34. [62 p.] (Out of print.) Simon, A. L. 1934. Wine and the wine trade. 2d ed. 129 p. Sir Isaac Pitman and Sons, London, England. United States Tariff Commission. 1939. Grapes, raisins, and wines. U. S. Tariff Comm. Rept. 2d ser. 134:1-408. Wine Advisory Board. 1940. The wines America likes. The wine handbook. 16 p. Quoted in : Wines and Vines 21(6): 5. Wine Institute. 1937-1941. Annual industry statistical surveys. (See especially the 5th, 21 p.) Wine Institute, San Francisco, Calif. (Mimeo.) References on Types and Composition of Wines Allen, H. W. 1931. The romance of wine. 264 p. E. P. Dutton and Co., New York, N. Y. 1933. Sherry. 117 p. Constable and Co., London, England. Belda, J. 1929. Vinos de Espana. 328 p. Compania Ibero-Americana de Publicaciones, Ma- drid, Spain. Biddle, A. J. D. 1900. The Madeira Islands. Vol. I. 324 p. Vol. II. 207 p. Hurst and Blachett, Lon- don, England. 207 See also the footnotes in each section of the text for additional references. 170 University of California — Experiment Station Brunet, E. 1927. Les vins de liqueur. 80 p. Librairie J.-B. Bailliere et Fils, Paris, France. GlANFORMAGGIO, F. 1910. Manuale pratico di vinifieazione, con appendice riguardante la fabbricazione dei vini Marsala, Moscato, Malvasia i Vermouth. 386 p. F. Battiatto, Can- tania, Italy. Griswold, F. G. 1929. Old Madeiras. 65 p. Duttons, Inc., New York, N. Y. Grossman, H. J. 1940. Grossman's guide to wines, spirits and beers. 404 p. Sherman and Spoerer, New York, N. Y. Herstein, K. M., and T. C. Gregory. 1935. Chemistry and technology of wines and liquors. 360 p. D. Van Nostrand Co., Inc., New York, N. Y. Hewitt, J. T. 1928. The chemistry of wine-making. [Great Britain] Empire Marketing Board Publication 7:1-57. 1939. The chemistry of wine-making. Science Progress 33:625-44. Kniphorst, L. C. E., and C. I. Kruisheer. 1937. Die Bestimmung von 2, 3-Butylenglykol, Acetylmethylcarbinol und Diacetyl in Wein und anderen Garungsprodukten. I. Entwicklung der Methodik. Zeit- schrift fur Untersuchungen der Lebensmittel 73:1-19. Mondini, S. 1900. II Marsala. 3d ed. Casa Editrice F. Ottavi, Casale Monferrato, Italy. Ottavi, O., and E. Garino-Canina. 1930. Vini di lusso. 8th ed. 424 p. Casa Editrice Fratelli Ottavi, Casale Monfer- rato, Italy. Prescott, S. C, and C. G. Dunn. 1940. Industrial microbiology. 541 p. McGraw-Hill Book Co., New York, N. Y. Pritzker, J. 1941a. Bericht liber die Sussweinkommission. Mitteilungen aus dem Gebiete der Lebensmitteluntersuchung und Hygiene 31:183—90. 1941o. Ueber die Zusammensetzung und Beurteilung von Malaga und Mistellen. Mitteilungen aus dem Gebiete der Lebensmitteluntersuchung und Hygiene 31:230-33. Schoonmaker, F., and T. Marvel. 1934. The complete wine book. 315 p. Simon and Schuster, New York, N. Y. 1941. American wines. 312 p. Duell, Sloan and Pearce, New York, N. Y. Semichon, L. 1911. Les vins de liqueur. Bevue de Viticulture 35:269-73, 300-6, 331-34, 364-70, 386-89. Shand, P. M. 1928. A book of French wines. 247 p. A. A. Knopf, New York, N. Y. 1929. A book of other wines than French. 185 p. A. A. Knopf, New York, N. Y. Simon, A. L., and E. Craig. 1933. Madeira. 153 p. Constable and Co., London, England. Bul. 651] Commercial Production of Dessert Wines 171 Winton, A. L., and K. B. Winton. 1935. The structure and composition of foods. Vol. IT. 904 p. John Wiley and Sons, New York, N. Y. References on Principles of Dessert- Wine Making Amerine, M. A., and M. A. Joslyn. 1940. Commercial production of table wines. California Agr. Exp. Sta. Bul. 639:1- 143. Brown, E. M. 1936. Stabilization of wine. Wines and Vines 17(8) : 12-13. Brown, E. M., and Victor de F. Henriques. 1935. Vinification in California wineries. Industrial and Engineering Chemistry, industrial edition 27:1235-40. Castella, F. de. 1925. Maturation of wine. Victoria Department of Agriculture Bul. 48:1-52. Cruess, W. V. 1934. The principles and practice of wine making. 212 p. The Avi Publishing Co., New York, N. Y. 1938. Commercial fruit and vegetable products. 2d ed. 798 p. (See p. 626-87.) McGraw-Hill Book Co., New York, N. Y. DUBAQUIE, J.-D. 1925. Oxydation et aeration des vins en futailles. Chimie et Industrie, Special no. p. 606-7. Gallagher, P. H., W. A. Stark, and P. Kolachov. 1941. The effect of various concentrations of ethyl alcohol on the fermentation rate of distillers' yeast. Journal of Bacteriology 41(1) :91-92. Harden, Arthur, 1932. Alcoholic fermentation. 243 p. (See especially p. 192-94.) Longmans, Green and Co., London, England. Joslyn, M. A. 1934. Possibilities and limitations of the artificial aging of wines. Fruit Products Journal 13:208, 241. 1935. Preliminary observations on the mellowing and stabilization of wine. Fruit Products Journal 15:10-12, 24. 1936. Process of aging or maturing wines. Food Industries 8:444-45, 449. Micksch, K. 1932. Development of aroma. [Translated title.] Brennerei Zeitung 49:185-86. Niehaus, C. J. G. 1930. Fortification of sweet wines. Farming in South Africa 4:511, 524. Theron, C. J., and C. J. G. Niehaus. 1938. Wine making. Union of South Africa Dept. of Agr. Bul. 191 : 1-98. (See espe- cially p. 58-69.) Twight, Edmund H. 1934. Sweet wine making in California. Parts I and II. California Grape Grower 15(10) :4-5; (11) :4-5, 7. Turbovsky, Morris. 1939. Technique in the standardization of wines. The Wine Review 7(1) :7-9, 34. 1937. Musts for sweet wine production. The Wine Review 5(9) : 12-14. 172 University of California — Experiment Station Ventre, Jules. 1933. Utilisation des marcs et des lies. Ecole Nationale d'Agriculture de Montpel- lier Annales 22:177-202. References on Winery Design and Operation Brunet, R. 1925. Manuel de tonnellerie. 284 p. Librairie J.-B. Bailliere et Fils, Paris, France. Grandchamp, L. 1935. Pour conserver les vins en futs. Revue de Viticulture 82:251-55. Hind, H. L. 1940. Brewing, vol. I. p. 1-506. vol. IT. p. 507-1120. John Wiley and Sons, New York, N. Y. JOSLYN, M. A. 1938. Protective bungs for barrels. The Wine Review 6(4) :16-17. Levine, S. 1937. Efficiency in water towers. The Wine Review 5(5) :24. 1938. Efficient winery operation. The Wine Review 6(5) :7-9. Lloyd, F. C. 1936. The art and technique of wine. 254 p. Constable and Company, London, Eng- land. Mondini, S. 1910. Costruzioni enotecniche. 251 p. U. Hoepli, Milano, Italy. Mrak, E. M., L. Cash, and D. C. Caudron. 1937. Effects of certain metals and alloys on claret- and sauterne-type wines made from vinifera grapes. Food Research 2(6) : 539-47. Mrak, E. M., D. C. Caudron, and L. M. Cash. 1937. Corrosion of metals by musts and wines. Food Research 2(5) : 439-54. O'Bert, L. R. 1938. Cooling tower efficiency. The Wine Review 6(4) : 14-15. Raphael, C. F. 1938. Planning the winery bottling room. The Wine Review 6(5) :17-19. Smith, C. 1940. The bottling room — winery stepchild. Wines and Vines 21(12) : 14-18. References on Red Dessert Wines Castella, F. de. 1908. Port. Victoria Department of Agriculture Journal 6:176-91. Ramires, A. B. 1929-31. Tratado de vinifacaQao. vol. I. 573 p. vol. II. 446 p. J. Rodrigues and Co., Lisbon, Portugal. Rocques, X. 1902. Les vins de Porto. Revue de Viticulture 18:9-11, 33-38, 201-16, 235-38. Simon, A. L. 1934. Port. 130 p. Constable and Co., London, England. Smyth, Alfred. 1900. Oporto et ses vins. 32 p. J.-B. Bailliere et Fils, Paris, France. Tait, G. M. 1936. Port from the vine to the glass. 174 p. Harper and Co., London, England. Bul. 651] Commercial Production of Dessert Wines 173 Twight, E. H., and M. A. Amerine. 1938. Port wine. Wines and Vines 19(2) :5-6. Vizetelly, Henry. 1880. Facts about Port and Madeira. 211 p. Ward, Lock and Co., London, England. References on Sweet White Dessert Wines Amerine, M. A., and A. J. Winkler. 1938. Angelica. Wines and Vines 19(9) :5, 24. Carpentieri, F. 1931. Enologia teorico-practica. 741 p. Casa Editrice F. Ottavi, Casale Monferrato, Italy. Helbig, W. A. 1938. What activated carbon is. The Wine Eeview 6(1) : 19-20, 24. Lachman, Henry. 1903. A monograph on the manufacture of wines in California. United States Dept. Agr. Bur. Chem. Bul. 72:25-40. Martinand, V. 1908. Les causes du pourvoir decolorant du charbon pur. Revue de Viticulture 30:569-72. Mathieu, L. 1935. Vins et eaux-de-vie a mauvais gouts. Revue de Viticulture 82:192. Mensio, C, and C. Forti. 1928. Enologia. 540 p. Union Tipografico — Editrice Torinese, Torino, Italy. Sannino, F. A. 1920. Trattato completo di enologia. 2d ed. vol. I. 482 p. vol. II. 368 p. V. Bona, Torino, Italy. Twight, E. H., and M. A. Amerine. 1938. The wines made from Muscat grapes. Wines and Vines 19(7) :3-4. References on Sherry and Other Rancio-Flavored Wines Amerine, M. A., and E. H. Twight. 1938. Sherry. Wines and Vines 19(5) :3-4. Anonymous. 1934. Madeira, the wine that never grows too old. California Grape Grower 15(6) : 18. California Agricultural Experiment Station. 1896. Report of the viticultural work during the seasons 1887-93, with data regard- ing the vintages of 1894-95. 466 p. (See especially p. 296-337.) State Print- ing Office, Sacramento, Calif. (Out of print.) Castella, F. de. 1909. Sherry: its making and rearing. Victoria Department of Agriculture Journal 7:442-46, 515-83, 621-30, 724-27. 1926. Sherry. Victoria Department of Agriculture Journal 24 : 690-98. 1929. Madeira. Victoria Department of Agriculture Journal 26:57-87; 27:651-60. Cruess, W. V. 1937. Lessons from Spanish sherries. The Wine Review 5(11) : 14-16, 36-37 ; (2) : 12-14,20; (5): 14-16. 174 University of California — Experiment Station Kickton, A., and O. Korn. 1924. Herstellung, Zusammensetzung und Beurteilung des Sherrys und seiner Ersatzweine. Zeitsclirift fiir Nahrungs- und Genussmittel 47:281-328. Marquis, H. H. 1936. California sherry production. The Wine Review 4(6) :6-7, 22. Niehaus, C. J. G. 1937. South African sherries. Farming in South Africa 12:82, 85. Roc que s, X. 1902. Le vin de Marsala. Revue de Viticulture 18 : 341-45, 371-77, 402-5. 1903. Les vins de liqueur d'Espagne. Revue de Viticulture 19:446-53, 501-5, 570- 73, 594-98. 1903. Les vins de Madere. Revue de Viticulture 19:11-18. 37-40. SCHANDERL, H. 1938. Die Nutzbarmachung des oxydativen Stadiums der Hefe bei der Trauben- und Beerenweinbereitung, sowie in der Brennereipraxis. Vorratspflege und Lebensmittelforschung 1:456-69. 1939. FruchtAveinbereitung nach alten und neuen Verfahren ( Sherry sierungsver- fahren) fiir Gewerbe und Hauschalt. Grunddlagen und Fortschritte im Garten- und Weinbau. Heft 53. 61 p. Verlag von E. Ulmer, Stuttgart-S., Germany. Thudichum, J. L. W., and A. Dupre. 1872. A treatise on the origin, nature and varieties of wine. 760 p. (See especially p. 632-71.) Macmillan and Co., London, England. Twight, E. H. 1936. California sherry wine making. Wines and Vines 17(4) :5, 15. Ref erences on Concentrate and Caramel-Sirup Production Barbet, e. 1932. Production de vins concentres les differentes manieres de les obtenir. Chimie et Industrie, Special no., p. 739-48. Brunet, R. 1934. Les mouts concentres de raisins. 128 p. Librairie J.-B. Bailliere et Fils, Paris, France. Castel, A. 1935. La fermentation des mouts concentres. Revue de Viticulture 82:395—98. Cruess, W. V. 1920. Commercial production of grape syrup. California Agr. Exp. Sta. Bui. 321: 399-416. (Out of print.) Cruess, W. V., and L. H. Hohl. 1937. Special methods of fermentation of sweet wines. Wines and Vines 18:8-9. Grignani, G. 1937. Composizione di concentrati di mosto di produzione siciliana. Annali di Chim- ica Applicata 27:209-12. Irish, John H. 1931. Fruit juice concentrates. California Agr. Exp. Sta. Bui. 392:1-20. (Out of print.) Mensio, C. . I mosti concentrati. Ediz. Ottavi, Casale Monferrato, Italy. Bul. 651 j Commercial Production of Dessert Wines 175 Boos, L. 1902. La concentration des vins, des mouts et des vendanges. 72 p. Feret et Fils, Bordeaux, France. 1925. La concentration appliquee au jus de raisin fermente ou frais et au raisin. Chimie et Industrie, Special no., p. 601-5. Venezia, M. 1938. Ricerche e considerazioni su alcuni succhi d'uva concentrati di produzione nazionale. Eegia Stazione Sperimentale di Viticoltura e di Enologia di Cone- gliano Annuario 8:215-39. References on the Production of Vermouth and Related Wines Arnou, L. 1905. Manuel du confiseur-liquoriste. 388 p. Librairie J.-B. Bailliere et Fils, Paris, France. BREVANS, J. DE. 1908. La fabrication des liqueurs. 3d ed. 568 p. (4th ed., 1920, 432 p.) Librairie J.-B. Bailliere et Fils, Paris, France. Cotone, D. A. 1922. II vino vermouth ed i suoi componenti. 2d ed. 546 p. Casa Editrice F. Mares- calchi, Casale Monf errato, Italy. Marescalchi, Arturo. 1934. Manuale dell'enologo e del canteniere. 9th ed. 208 p. Casa Editrice F. Mares- calchi, Casale Monf errato, Italy. Mattiralo, O. 1915. Sulla coltivazione e sul valore delle "Artemesia" usate nella fabricazione dei Vermouths. Regia Academia d'Agricoltura di Torino Annali 58:225-77. Maurizio, A. L. 1933. Geschichte der gegoren Gatranke. 262 p. Paul Parey, Berlin, Germany. Ottavio, Ottavi, and E. Garino-Canina. 1930. Vini di lusso. 8th ed. 424 p. Casa Editrice Fratelli Ottavi, Casale Monfer- rato, Italy. Sebastian, Victor. 1909. Traite pratique de la preparation des vins de luxe. 656 p. Masson et Cie., Paris, France. Strucchi, A., and F. Carpentieri. . II Vermut. 3d ed. Casa Editrice F. Ottavi, Casale Monferrato, Italy. Vandone, Edoardo. 1930. Manuale pel liquorista. 323 p. Casa Editrice F. Marescalchi, Casale Mon- ferrato, Italy. References on the Clarification and Stabilization of Wines Anonymous. 1937a. The use of diatomaceous filter-aids. The Wine Beview 5(5) :15-17. 1937&. Filtration with diatomaceous silica. The Wine Eeview 5(8) : 19, 30. Dubaquie, J. 1926. Le chauff age de la vendange et le development des vin en moelleux et en bouquet. Academie d' Agriculture de France Comptes Rendus 12:52-53. Dern, K. L. 1940. Filtration with diatomaceous silica. Wines and Vines 21(4) : 23-24. 176 University of California — Experiment Station Genin, g. 1934. La filtration industrielle. 446 p. Dunod, Paris, France. Hartman, E. J. 1939. Colloid chemistry. 556 p. Houghton Mifflin Co., New York, N. Y. Kenney, J. 1940. Modern wine filtration. Wines and Vines 21(9) : 20-22; (10) : 20-22. Krause, C. V. 1940. Filtration. Wines and Vines 21(1) : 18-19. Laborde, J. 1919. Recherches sur le viellissement du vin. Revue de Viticulture 48:225-30, 241- 44, 386-90; 49:38-41, 49-51, 65-69. Mathieu, L. 1925. Theorie et pratique du collage des vins. Chimie et Industrie, Special no., p. 608-14. 1934. Sur la refrigeration industrielle des vins. Revue de Viticulture 81:347-49. Meissner, R. 1920. Technische Betriebskontrolle im Weinfach. 538 p. E. Ulmer, Stuttgart, Germany. Ribereau-Gayon, J. 1936. Notes sur le vieillissement des vins. Societe des Sciences Physiques et Natur- elles de Bordeaux Proces-verbeaux des Seances 1936-37:26-28. 1939. Les phenomenes colloidaux dans le vin. Revue de Viticulture 90:255-64, 275-87. References on the Preparation of Wines for Market Baker, W. F. 1940. Wine bottles. Wines and Vines 21(5) :13. Gray, P. P., I. Stone, and H. Rothchild. 1941. The action of sunlight on beer. Wallerstein Laboratories Communications 4(ll):29-40. Herstein, K. M., and T. C. Gregory. 1935. Chemistry and technology of wines and liquors. 360 p. D. Van Nostrand Co., Inc., New York, N. Y. Rosenbloom, M. V., and A. B. Greenleaf. 1940. Bottling for profit, a treatise on liquor and allied industries. 192 p. American Industries Surveys, New York, N. Y. Samuel, C. 1940. The story of cork. Wines and Vines 20(11) : 20-21. United States Bottlers Machinery Co. 1940. Bottling engineer handbook. 191 p. United States Bottlers Machinery Co., New York, N. Y. References on Bacterial Diseases and Other Disorders of Wine Baillot d'Estivaux, M. L. 1935. Sur un developpement intense du microbe de la tourne dans un milieu tres alcoolique. Societe des Sciences Physiques et Naturelles de Bordeaux Proces- verbaux des Seances 1934-35:35-42, 54-56. Cettolini, S. 1908. Malattie, alterazioni e difetti del vino. 379 p. U. Hoepli, Milano, Italy. Bul. 651] Commercial Production op Dessert Wines 177 Cruess, W. V. 1937. Observations of '36 season on volatile acid formation in Muscat fermenta- tions. Fruit Products Journal 16:198-200, 215, 219. 1938. Non-bacterial spoiling of wine. Wines and Vines 19(1) : 20-22. Couche, D. D. 1935. Modern detection and treatment of wine diseases and defects. 98 p. The Technical Press, Ltd., London, England. Douglas, H. C. 1938. Microbiological problems of the sweet wine industry. Wines and Vines 19 (7): 16-17. Douglas, H. C, and L. S. McClung. 1937. Characteristics of an organism causing spoilage in fortified sweet wines. Food Eesearch 2:471-76. Fornachon, J. C. M. 1938. Bacterial fermentations in fortified wines. Australian Wine Board Publica- tion on Wine Investigations. 19 p. (Mimeo.) Garino-Canina, Ettore. 1935. II potenziale di ossido riduzione e la tecnica enologica. Annali di Chimica Applicata 25:209-17. Geloso, Jean. 1931. Relation entre le vieillissement des vins et leur potentiel d'oxydo-reduction. Annales de la brasserie et de la distillerie 29:177-181, 193, 197, 257-61, 273-77. (See also: Chimie et Industrie 27:430-31. 1932.) Marsh, G. L. 1940. Metals in wine. The Wine Eeview 8(9) :12-13, 24; (10) :24-28, 29. Martin, E. 1925. Determination chimique du degre alcoolique des vins et d'une maniere generale, de tout liquide renfermant de l'alcool. Chimie et Industrie, Special no., p. 589-91. Martin, R., and M. Castaing. 1934. La casse ferrique des vins blanc et son traitement rationnel. Revue de Viti- culture 81 : 235-38. Mathieu, L. 1934. Elimination du cuivre accidental des mouts et des vins. Revue de Viticulture 81:251-53. Niehaus, C J. G. 1932. Mannitic bacteria in South African sweet wines. Farming in South Africa 6:443-44. Ribereau-Gayon, J. 1933. Contribution a l'etude des oxydations et reductions dans les vins. 2d ed. 214 p. Librairie Delmas, Bordeaux, France. 1939. Determination du potentiel d'oxydo-reduction des vins. Annales des Falsifi- cations et des Fraudes 32:385-400. English translation in: The Wine Review 8(6):16-18; (9) : 16-17, 26-27. Shimwell, J. L. 1941. The lactic acid bacteria of beer. Wallerstein Laboratories Communications 4(11) :41-48. 178 University of California — Experiment Station References on the Analysis of Wines Association of Official Agricultural Chemists. 1940. Official and tentative methods of analysis. 5th ed. 757 p. Association of Official Agricultural Chemists, Washington, D. C. Bames, E., B. Bleyer, G. Buttner, W. Diemair, H. Holthofer, O. Reichard, and E. Vogt. 1938. Alkoholische Genussmittel. Vol. 7. 828 p. In: Bomer, A., A. Juckenack, and J. Tillmans. Handbuch der Lebensmittelchemie. Julius Springer, Berlin, Germany. Bridges, M. A. 1935. Food and beverage analyses. 246 p. Lea and Febiger, Philadelphia, Pa. Cruess, W. V., M. A. Joslyn, and L. G. Saywell. 1934. Laboratory examination of wines and other fermented fruit products. Ill p. The Avi Publishing Co., New York, N. Y. Dujardin, J., and L. and R. Dujardin. 1928. Notice sur les instruments de precision appliques a Poenologie. 6th ed. 1,096 p. Dujardin-Salleron, Paris, France. Dutoit, Paul, and Marcel Duboux. 1912. L'analyse des vins par volumetrie physico-chimique. 189 p. F. Rouge et Cie., Lausanne, Switzerland. Emiliani, E. 1938. La determinazione ebulliometrica del grado alcoolico nei vini dolci. Annali di Chimica Applicata 28 : 409-12. Eynon, Lewis. 1923. Wines and potable spirits. Vol. I, p. 221-65. In: Allen's commercial organic analysis. P. Blakiston's Sons & Co., Inc., Philadelphia, Pa. Fabre, J.-Henri. 1936. Procedes modernes de vinification. II. Analyse des vins et interpretation des resultats analytique. 2d ed. 346 p. La Typo-Litho et J. Carbonel Reunies, Algeria. Fortune, W. B., and M. G. Mellon. 1938. Determination of iron with o-phenanthroline. Industrial and Engineering Chemistry, analytical edition 10(2) : 60-64. Hodgman, C. D. 1940. Handbook of chemistry and physics. 23d ed. 2,221 p. Chemical Rubber Publishing Co., Cleveland, Ohio. Jacobs, Morris B. 1938. The chemical analysis of foods and food products. 537 p. (See especially p. 399-411.) D. Van Nostrand Co., Inc., New York, N. Y. Jaulmes, P., and P. Espezel. 1935. Le dosage de l'acetaldehyde dans les vins et les spiritueux. Annales des Falsifications et des Fraudes 28 : 325-35. Joslyn, M. A. 1938-40. Report on volatile acids in wine. Association of Official Agricultural Chemists Journal 21:166-74; 22:210-20; 23:183-189. Bul. 651] Commercial Production of Dessert Wines 179 Joslyn, M. A., and C. L. Comar. 1938. Determination of acetaldehyde in wine. Industrial and Engineering Chem- istry, analytical edition 10(7) : 364-66. Marsh, G. L., and K. Nobusada. 1938. Iron determination methods. The Wine Review 6(^9-) : 20-21. Paronetto, L. 1938. La determinazione delPalcole nei vini liquorosi c vermut per ossidazione cromica. Annali di Chimica Applicata 28:164-69. Procopio, Mario 1939. II grado ebulliometrico dei vini dolci. Annali di Chimica Applicata 29 : 74-77. Saywell, L. G., and B. B. Cunningham. 1937. Determination of iron. Industrial and Engineering Chemistry, analytical edition 9(2): 67-69. Tarantola, C. 1934. Determinazione delle aldeidi nel vino con il fotometrodi Pulfrich e con il colorimetro fotoelettrico. Annali di Chimica Applicata 24:615-25. Tillmans, J. 1927. Lehrbuch der Lebensmittelchemie. 387 p. Julius Springer, Berlin, Germany. Windisch, Karl. 1896. Die chemische Untersuchung und Beurtheilung des Weines. 351 p. Julius Springer, Berlin, Germany. INDEX acetal, 19, 60, 101 acetadehyde, 19, 61, 89, 90, 97, 101, 107; see also aldehydes acetic acid, 13, 14, 19, 22, 23, 43, 44, 59, 98, 143, 146; bacteria, 13, 41, 57, 96, 102, 143- 44; determination of, 156; in Angelica, 25; in "aromatic" wines, 34; in Madeira, 32; in Malaga, 30 ; in Marsala, 32 ; in muscatel, 28; in port, 22, 23; in sherry, 30, 94; in Tokay, 32; in vermouth, 34, 124; in white port, 27; influence of sulfur dioxide on, 41, 42 ; legal limits, 14 ; limit of error in deter- mination of, 167; reduction by yeast, 98, 99 Acetobacter, 144 ; see also bacteria acetylmethylcarbinol, 18 acidity, 43, 48, 90, 105, 135, 167; correction of, 44, 136, 146; determination of, 156; in- fluence of climate on, 13, 35; influence on fermentation, 48; influence on taste, 13; limit of error in the determination of, 167; of Angelica, 24, 25; of "aromatic" wines, 34; of Madeira, 32; of Malaga, 30; of Mar- sala, 32; of muscatel, 28; of musts, 13, 14, 35-38, 40, 74, 93; of port, 22, 23; of sherry, 30, 94, 99, 106; of Tokay, 32; of white port, 27; of vermouth, 34, 117, 119, 124; range in, 13, 14; relation to pH, 13, 36, 93; relation to spoilage, 44, 143, 146; total, 22, 23, 156, 167; see also acetic, citric, gallic, humic, levulinic, malic, succinic, and tartaric acids, and tartrates acknowledgment, 168 acreage, 8, 9, 26 adulteration, 17 aeration, 54, 99; during sherry heating, 106; in centrifuging, 133; of must, 110; see also oxidation aging, 17, 41, 58-61, 63, 78, 80-81, 87-88, 96, 127, 144; changes during, 20, 22; in- fluence of fortification on, 48 ; of blends, 136; of brandy, 105; of concentrate, 113; of sherry, 15, 29, 96, 100, 101, 105, 108; of vermouth, 124, 125; over-, 26, 102; quick, 60, 61, 80, 88; relation to size of container, 67, 68, 87; time for, 4, 13, 60, 81, 87, 108, 110 albariza soil, 92 albumin, 131 ; see also proteins alcohol, amyl, 14; butyl, 14; changes in during aging, 15, 60, 101; determination of, 51, 79, 81, 150-52, 154-56; during fortification, 51—56; effect of Acetobacter on, 144; effect on tasting, 139; ethyl, 14; heptyl, 14; hexyl, 14; higher, 14, 57, 58, 139; in "aromatic" wines, 34; in Angelica, 25, 81; in blends, 137 ; in Madeira, 32 ; in Malaga, 30 ; in Mar- sala, 32 ; in muscatel, 26, 28 ; in port, 22, 23, 73, 78 ; in sherry, 29, 30, 94, 95, 99, 102 ; in Tokay, 32; in vermouth, 34, 117, 119, 124; in white port, 27; influence on Balling reading, 45, 46 ; influence on composition, 12, 15; influence on expansion, 139, 140; isoamyl, 14 ; isobutyl, 14 ; limits, 14, 46, 47, 48, 73, 81, 87, 91, 92, 104, 110, 115; limit of error in determination of, 167; losses, 15; methyl, 14, 15; n-butyl, 14; n-propyl, 14; oxidation of, 88, 90, 98, 148; table for correcting, 153; terpenic, 117; tolerance of bacteria for, 80, 144, 145, 146 ; tolerance of yeasts for, 47, 48, 99, 101, 146; use of concentrate to increase, 111; vapors, 73, 106; see also fortification and fortifying brandy aldehydes, 42, 60, 139, 148, 149; determina- tion of, 160-61; in Angelica, 25; in fortify- ing brandy, 57, 58, 139 ; in muscatel, 28 ; in port, 23; in sherry, 19, 24, 29, 30, 99, 107; in vermouth, 117, 118; in white port, 27, 88 ; limit of error in the determination of, 167; paraldehyde, 115; see also acetalde- hyde Aleatico, 39; composition of, 36; see also mus- catel alkaloids, in herbs for vermouth, 117, 118 Alicante Bouschet, 39, 40 ; acreage, 9 Alicante Ganzin, 30, 40 allspice, 120, 123 aloe, 118, 120, 123 aluminum, for coils, 106; in bottles, 141; in clays, 132 Alvarelhao, 39 amontillado, 90, 91, 92, 94, 100, 101 analyses, 107, 138, 150-68; alcohol, 150-56; aldehydes, 160-61; 2, 3-butylene glycol, 167; color, 161-62; coloring matter, 160; concentrate, 113; copper, 163—66; esters, 161; glycerin, 167; hydroxymethylfurfural, 166-67; interpretation of, 167-68; iron, 162—63; reducing sugars, 158—60; room for, 62, 63; sulfur dioxide, 167; tannin, 160; time for, 59, 103; total acid, 156; volatile acid, 156 angelica (herb), 117, 118, 121, 123 Angelica (wine), 21, 22, 48, 54, 55, 59, 81, 82-89, 109, 112, 135, 145; composition, 24, 25 ; composition of musts for, 35 ; for blending, 31; production of, 6, 8; use in making California Tokay, 111; use in mak- ing vermouth, 124; varieties for, 36, 37 angostura, 118, 120, 123, 125 anise, 118, 120, 123; star, 118, 121, 123 anthocyanin, 20, 101, 147, 148, 149; see also color antiseptics, 71, 72, 73, 141 appetizer wines, 3, 24, 33, 108, 125 Armenian wine, 96 aroma, 14, 18, 19, 20, 36, 42, 57, 58, 60, 61, 78, 92, 136, 138, 139 "aromatic" wines, 33, 34, 125 arrope, 95 asbestos, for filtering, 20, 130 ash, 89, 93, 94, 128; alkalinity of, 59 asoleo, 95 asphyxiation, 76 Australia, 29, 101, 102 bacteria, 42, 44, 57, 67, 71, 72, 74, 80, 82, 99, 125, 143-44; acetic acid, 13, 41, 57, 96, 102, 143-44; clouding due to, 126, 145; esters formed by, 19; hair bacillus, 144-46; lactic acid, 48, 57, 58, 142, 143, 144, 146; mannite formed by, 14, 144, 146 ; see also pasteurization and yeasts Balling 14, 39, 54, 74, 78, 83, 92, 97, 138; chart for determining time to fortify from, 49 (fig. 1) ; hydrometer, 45-46, 157; influ- ence on amount of sulfur dioxide to use, 42 ; influence on period of fermentation, 41 ; limit of error in the determination of, 167; of Angelica, 24, 25, 48; of concentrate, 112 113; of Madeira, 32; of muscatel, 26, 28 48 ; of musts, 35-38, 40, 95 ; of port, 22, 23 48 ; of sherry, 30, 48 ; of Tokay, 32 ; of white port, 27; table for correcting readings, 157 table of specific gravity corresponding to readings, 158; to fortify, 49-55, 78, 85 see also sugar Balling-acid ratio, 35-38, 40 barrels, see containers Bastardo, 39 ; see also Trousseau Baume, see Balling beer and breweries, 60, 76, 147 beet sugar, see sucrose bentonite, 80, 88, 89, 124, 125, 127, 131-32; composition of, 132; see also Spanish earth benzoin, 120 Black Prince, 37 blending, 50, 63, 96, 99, 100, 105, 125, 134- 38, 139; for California Tokay, 110; for Madeira, 109; for Marsala, 110; formulas for, 51-53, 136-38; of herbs for vermouth, 119, 122; of ports, 39; purpose of, 134; [181] 182 University of California — Experiment Station technique, 136—38; use of Angelica for, 24; use of reduced musts, 33; see also solera blood, 110, 130, 132 Boal di Madeira, 38 bodega, 100 body, of Marsala, 109; of wines for vermouth, 117; see also extract Boletus laricus, 121, 123 Botrytis cinerea, 111 bottles, 140, 141 (fig. 13); breakage due to freezing, 140 ; capsuling, 142 ; closures for, 142; headspace in, 142; labeling of, 142; sediment in, 22, 126, 139, 141, 145, 147 bottling, 24, 73, 125, 135, 136, 139-43; fillers for, 140, 141—42; foaming during, 130, 141; room, 62, 63, 64, 70 (fig. 10), 141 ; washers, 71 bouquet, see aroma brandy, 4; in vermouth making, 119; room, 62, 63, 64; taste, 58, 87, 107; see also fortifying brandy "break" test, 107, 108 Brix, see Balling buffer capacity, 14 Bureau of Internal Revenue, 4, 6, 8, 46, 116, 142; Alcohol Tax Unit of, 56, 61; see also gauger, legal restrictions, and taxes Burger, 38 ; acreage, 9 butt, sherry, 68, 93, 95, 98, 100, 101, 102, 134 butylene glycol, 18; determination of, 167 Byrrh, 125 calamus, 117, 118, 120, 123 calcium, from filter pads and concrete tanks, 147; sulfate, 93; tartrate, 126, 147; see also ash California, acreage of grapes in, 8, 9, 26; De- partment of Public Health, 14, 46 ; dry sherry, 14, 19, 21, 30, 31, 109; Madeira, 31, 32, 109, 114; Malaga, 31, 108; Marsala, 31, 32, 110, 114; port, 21-24, 31, 48, 54 (table 23), 55 (table 24), 73-81, 135; pro- duction of wine in, 5, 10 ; sherry, 14, 17, 19, 21, 29, 30, 31, 33, 37, 38, 48, 54, 55, 59, 90, 111; sweet sherry, 14, 21, 31, 109, 111; tawny port, 21 ; Tokay, 14, 21, 31, 32, 33, 110, 111; white port, 21, 25, 26, 27, 55, 81, 83; wine types, 21—33; see also Angelica, muscatel, and vermouth camomile, 17; Roman, 118, 121, 123 (foot- notes) cane sugar, see sucrose cap, 41, 57, 74; management of, 75-77; sub- merged, 75, 76 capsuling, 142 caramel, 17, 111, 114-15, 147; flavor, 21; solubility of, 114; use of, 115, 119 caramelization, 4, 33, 59, 103, 113, 114, 138 carbon, dioxide, 54, 60, 76, 98; monoxide, 73 cardamon, lesser, 118, 119, 121, 123 Carignane, 39, 40, 82, 143; acreage, 9; for California Marsala, 110 cascarilla, 118, 120, 123 casein, 131, 132 cask, see containers cask borer, 73 casse, copper, 149; iron, 149; oxidasic, 150 Catawba, 81 catechu, 124 cellar, for sherry, 108; storage, 56, 58, 61, 62, 63, 64, 67, 68, 79, 80, 81, 87, 100, 102, 125 cellulose, cap, 142 ; for filtering, 20 centaury, European, 117, 118, 120, 123 Central Valley, production in, 10 centrifuging, 15, 133 champagne, 60 charcoal, 24, 26, 31, 82, 132, 139, 149; use in preparing white port, 88—89 Chateau Chalon, 97, 99 chincona, 117, 118, 120, 122, 123, 125 cinnamon, 118, 119, 120, 122, 123, 125 Cinsaut, 40 citric acid, 13, 44, 110, 111, 117, 124, 125, 149 ; for washing filter pads, 130 clarification, see filtration, fining, and settling climate, influence on composition, 35 closures, 141, 142 cloudiness, 59, 61, 75, 85, 87, 105, 108, 111, 117, 125, 126, 127, 128, 132, 136, 139, 141, 145, 146, 147, 150; due to metals, 126, 147, 149 ; of vermouth, 117, 124 clove, 118, 119, 120, 123, 125 "Coagol," 131 coca, 118, 120, 123 cocktail, 108, 116; use of vermouth in, 33; see also appetizer wines coils, metals for, 106, 149 colloids, in caramel, 115; influence on filtra- tion, 129; precipitation of, 148; protective, 127, 131, 149; removal, 59, 129, 133; see also gums and pectins color, 22, 26, 41, 59, 73, 81, 82, 84, 88, 92, 95, 98, 111, 117, 128, 139; blending to correct, 135; darkening of, in reduced musts, 114; extraction of, 41, 75, 77; of Angelica, 24, 135, 139; of caramel, 114, 115; of concentrate, 113; of Italian ver- mouth, 117, 119; of Marsala, 109, 110; of muscatel, 26, 135, 139; of sherry, 31, 33, 107; of Tokay, 111, 139; of varieties for port, 40; of white port, 24, 88-89; pre- cipitation of, 60, 126, 147; produced by heating, 31, 33, 107; standardization of, 161—62; see also anthocyanin and port coloring matter, analysis for, 160 composition, see Angelica, Madeira, Malaga, Marsala, muscatel, port, sherry, Tokay, white port, and vermouth compressed air, 51, 79, 87 concentrate, 15, 16, 33, 40, 91, 97, 110, 119, 136, 139; addition to must, 114; composi- tion of, 115; methods for producing, 113; vacuum pans for, 112; see also reduced musts Concord, 73, 106 concrete, see containers congenerics, 57, 58 consumption, 6, 7 containers, 31, 48, 58, 87, 109, 134; care of, 71, 72, 73; concrete, 62, 63, 64, 66, 67 (fig. 8), 68, 69, 70, 71, 80, 113; condi- tioning, 68, 69; filling of, 87, 88; head- space in, 139—40; influence on condition of wine, 15, 59, 60, 90, 103, 125, 131; oak, 22, 29, 59, 62, 63, 68, 69, 73, 80, 87, 100, 101, 102, 103, 104, 105, 113, 114, 134. 135, 140; pure culture, 65, 66; redwood, 50 (fig. 2), 62, 63, 67, 68, 69, 80; size for bulk quality, 64, 65; storage, 62, 63, 64, 67, 68 ; see also butt and pipe contraction, after fortification, 51 (table 21), 52, 53, 54; due to temperature, 139, 140 conveyors, 62, 67 cooper, 63, 64, 72 copper, 106, 113; clouding due to, 147, 149; determination of, 163—66; see also metal Cordova, 92 coriander, 117, 118, 120, 123, 125 corks, 142, 146 cortado, 94 "cottony mold," see bacteria cream of tartar, see tartrates crushing, 41, 62, 64, 65, 74, 75, 82, 83, 108, 110 Davis, 35, 36-38, 40 deposits, amorphous, 147-48; of copper casse, 149; of iron casse, 149; of oxidasic casse, 150; see also filtration, fining, and settling dessert wine, consumption, 7, 8 ; definition, 3 ; difference from table wines, 12 ; directions for making, 73—111, 115—25; principles of making, 33—61; production, 5, 6, 8; pro- duction by districts in California, 10 ; types, 21-33, 34; see also red dessert wine, white dessert wine dextrose, 15, 16, 22, 33, 40, 44, 98, 112; sweetness of, 16; see also sugars Bul. 651] Commercial Production op Dessert Wines 183 diacetyl, 18 diatomaceous earth, see filter aids diffusion battery, 56, 65 distillery, 62 (fig. 3), 63 (fig. 4) ; see also still distilling material, 44, 65, 69, 78, 83, 144; muscat, 83; production of, 56, 57 dittany of Crete, 120 Douro district, 22 dry dessert wine, white, see sherry dry vermouth, see vermouth (dry) Dubonnet, 125 ebullioscope, 51, 150, 151 ecclesiastical wine, 14, 46 egg white, 130 Eighteenth Amendment, see Prohibition elder, flower, 117, 120, 123 elecampane, 118, 120, 123 Emperor, 11 enzymes, 58, 59, 82 Erbalus di Caluso, 37 essential oils, in vermouth, 34, 122 esters, 15, 19, 20, 57, 58, 59, 60, 61, 97, 101, 107, 110; determination of, 161; in herbs for vermouth, 117; in muscatels, 28; in port, 23; in sherry, 30, 94; limit of error in determination of, 167 es tufas, 109 ethyl acetate, see esters ethyl alcohol, see alcohol extract, 37, 45, 59, 91, 107, 128; determina- tion of, 157—58; in Angelica, 25; in "aro- matic" wines, 34; in Madeira, 32; in Ma- laga, 30; in Marsala, 32; in muscatel, 26, 28; in port, 22, 23; in sherry, 30, 94; in Tokay, 32 ; in vermouth, 34 ; in white port, 26, 27; limit of error in determination of, 167 Feher Szagos, 38 fennel, 118, 120, 123 fenugreek, 118, 120 fermentation, 38-46, 74, 75, 78, 83-84, 108, 111, 112 ; for sherry, 93, 102, 103 ; influence of acidity on, 13; influence of alcohol on, 47, 48 ; influence of period of on composi- tion, 41, 43, 44; influence of temperature on, 41 ; influence on flavor, 39 ; of concen- trate, 33; products of, 12, 13, 14, 15, 18, 19, 43-44, 48, 84, 96, 97; rate of for levu- lose and dextrose, 15; spoilage during, 41, 75, 143; stuck, 37, 41, 45, 75, 77, 144 fermenters, 62, 63, 64, 65, 66 (fig. 6), 67, 84, 105 ; larger type, 93 ; size of, 65 ; submerged cap, 74, 76 fermenting room, 62, 63, 64, 65, 66 (fig. 6), 67 filling, 87-88, 134; of sherry butts, 100, 103; to prevent overoxidation, 58 film, see yeast filter aids, 89, 129, 130, 139, 147; new ma- terials for, 130 filters, 105; candle, 130; pad, 70, 130; plate and frame, 70, 130 ; polishing, 70, 130, 140 ; presses, 89, 129 filtration, 15, 20, 60, 80, 88, 103, 105, 108, 110, 127, 128, 129-30, 146; of vermouth, 124; sterilization, 126; testing efficiency of, 130 ; versus fining, 132-33 fining, 20, 60, 80, 103, 105, 108, 127, 130-33; oxen's blood for, 110, 132; of vermouth, 124; versus filtration, 132-33; see also bentonite fi.no, 90, 91, 92, 94, 95, 96, 98, 100, 101, 102, 103, 107 Flame Tokay, 31, 38, 143 flavor, 18, 24, 26, 29, 37, 38, 39, 40 (table), 43, 44, 45, 49, 57, 58, 75, 81, 87, 91, 92, 111, 112, 114, 126, 130; brandy, 57, 58, 79, 87, 107; caramel, 21, 83, 90; charcoal, 26; cooked, 21, 112; extraction of, 41, 75, 76, 77, 84-85; foxy, 106; from filters, 130; hvdroxymethylfurfural, 17; muscat, 36, 37, 39, 81, 119; oak, 20, 59, 60, 61, 68, 91, 101, 103, 104, 105, 108; off-, 26, 42, 48, 57, 83, 89, 106, 112, 130, 132, 136, 142, 144; rancio, 4, 21, 24, 33, 59, 81, 90, 91, 103, 104, 108, 109, 111; sherry, 24, 29, 38, 59, 90, 92, 93, 95, 96, 97, 100, 101, 107; tannin, 20; vermouth, 116, 117, 118, 119, 120, 122, 124, 125; see also tasting and wood floors, design of, 71 flor, 29, 98, 99, 101, 102, 103 "flowering," see yeast foaming, 60, 130, 141 formula for blending, 137, 138; for fortifying, 51-54 fortification, 4, 14, 24, 26, 29, 33, 39, 40, 42, 46-55, 78-80, 81, 82, 85, 87, 95, 96, 102, 109, 110, 111, 125, 135, 136; changes in composition during, 54—55 ; contraction on, 51, 52, 53, 54; chart for determining time for, 49 (fig. 1); dilution effect, 13, 14, 18, 57; heat evolved, by, 54, influence on levu- lose-dextrose ratio, 15 ; of vermouth, 116, 119, 122, 124; procedure, 50, 51, 79, 80 fortified wine, 3, 15, 81, 109; see also Angel- ica, dessert wine, Madeira, Malaga, Marsala, muscatel, port, sherry, Tokay, and white port fortifying brandy, 4, 11, 12, 14, 40, 42, 44, 48-51, 59, 61, 81, 96, 102, 104, 105, 108, 110, 116; addition by installments, 48, 102, 135; calculation of amount required, 51— 54; measurements of, 50, 51, 53, 79; mix- ing, 51, 54, 79, 87; proof of, 57, 58, 79, 81, 109; quality of, 49, 78, 79, 139 fortifying room, 50 (fig. 2), 56, 62 (fig. 3), 63, 64 (fig. 5), 79, 80, 84, 85 fortifying tank, 50 (fig. 2), 51, 56, 79, 80 foxy flavor, removal of, 106 France, 3, 24, 26, 87, 97, 99, 115, 120, 121, 123, 124, 125 fraxinella, 120 French vermouth, see vermouth (dry) Fresno, 37, 38, 42, 55; "mold," 144; produc- tion in, 8, 10 Frontignan, 26, 87 fruit flies, 67 fungi, 47, 72, 73 ; see also yeast Furmint, 31 fusel oils, 57, 58, 139 galingale, 118, 120, 123 gallic acid, 148 gallotannin, 148 gauger, United States, 4, 50, 51, 79, 85; office for, 51 ; official samples of, 54, 56 gelatin, for fining, 20, 80, 131; wine overfined with, 132 gentian, 117, 118, 120, 123; wine, 125 germander, 120, 123 Germany, 3, 89, 132 ginger, 120 glucosides, in herbs for vermouth, 117, 118 glycerin, 18, 19, 43; determination of, 167; in Angelica, 25; in "aromatic" wines, 34; in Madeira, 32 ; in Malaga, 30 ; in Marsala, 32 ; in muscatel, 28; in port, 23; in sherry, 30; in Tokay, 32; in vermouth, 34; in white port, 27 gondola trucks, 82 gout de ranee, 90 ; see also flavor (rancio) Grand Noir, 39, 40 grapes, composition for dessert wines, 33, 35— 39, 40 ; cost of production, 8 ; raisin, 8, 9, 11, 13; table, 1, 8, 9, 11, 13; varieties for dessert wines, 33-39, 40; wine, 9, 11; yield per acre, 8 ; see also crushing, harvesting, pressing, and raisining Greece, 26, 28 Green Hungarian, 38 Grenache, 37, 38, 82; acreage, 9 Grillo, 37 gums, 56, 85, 115, 148 ; in herbs for vermouth, 117 gypsum, 93, 102 184 University of California — Experiment Station hair bacillus, 144-40 Hansenula saturnus, 97 hart's tongue, 120 harvesting, 26, 35, 82, 110; date of, 36-38, 40 ; for port, 74 ; for white dessert wines, 82, 83 health, of workers, 73, 76, 106 heat exchanger, see pasteurization and refrig- eration heating, 13, 17, 18, 31, 33, 85, 103-08, 131, 139; aeration during, for sherry, 106; dis- solves tartrates, 128; electrical, 61, 80, 107; in making muscatel, 84—85 ; in making port, 76, 77, 80; in stabilizing wine, 126, 127, 133 ; of Madeira, 108, 109 ; of Malaga, 108 ; of muscatel, 26 ; of sherry, 29 ; of sugar, 114—15; of Tokay, 111; of vacuum pans, 113; over-, 75, 77, 112; temperature for Madeira, 109; temperature for sherry, 103, 105, 106; various methods, 103, 105, 106, 107 herbs, 3, 11, 33, 117-22; amounts of, to use, 122, 125; common commercial, scientific, Italian and French names of, 120, 121 ; con- stituents in, 117, 118; flavor of, 115, 122; for dry vermouth, 123, 124; masceration of 116, 119, 125; medicinal properties of, 117; possibility of growing in America, 122; use in Marsala, 110 higher alcohols, see fusel oils hogsheads, see containers (oak) homemade wine, production, 5, 6 hop, 118, 121, 123 horehound, common, 120 hose, rubber, 70 humic acid, 113 humidity, of storage room, 60, 71, 73, 100 humin, in caramel, 115 Hungary, 31, 32 hydroxymethylfurfural, 17; determination of, 166-67 hypochlorite, 71, 72, 73 hyssop, common, 118, 120, 123 imported wine, amount, 7, 11; vermouth, 12 invert sugar, sweetness of, 16 (footnote 23) Inzolia bianca, 38 iron, clouding due to, 147, 149; determination of, 162—63 ; see also metal isinglass, 131 isosaccharosan, in caramel, 115 Italian vermouth, see vermouth (sweet) Italy, 3, 24, 28, 32, 89, 115, 117, 119, 120, 121, 122 Ives Seedling, 73 Jerez de la Frontera, see Spain kaolin, 131, 132 labeling, 142-43 ; see also legal restrictions and nomenclature lactic acid bacteria, 48, 57, 58, 142, 143, 144, 146 lagar, 93 Lavulosins, 18 leaguer, 102 lees, 56, 59, 60, 81, 96, 98, 102, 103, 108; brandy, 80 legal restrictions, 14, 31, 40, 47, 48, 49, 50, 77, 79, 82, 111, 135, 142, 143 ; for muscatel, 26; in Portugal, 22; vermouth, 116 lemon balm, 118, 121, 123 levulinic acid, 18 levulose, 15, 33, 146 ; levulose-dextrose ratio, 15, 16, 22, 44; sweetness of, 16 licorice, 124 Likorweine, 3 lime, 92 ; see also antiseptics Lodi, 35, 37, 38, 55 ; dessert wine in, 8, 9, 10 lungwort, 121, 123; lichen, 121 Madeira, 26, 31, 59, 90, 108-09, 111; com- positions, 32 ; esters in, 20 ; production of, 6, 8; sucrose in, 17 Malaga (grape), 38 Malaga (wine), 16, 31, 38, 81, 90, 91, 108, 111, 145; composition, 30; production of, 6 malic acid, 13, 19, 111 Malmsey, 26, 38 Malvasia bianca, 36 mannite, 15, 146, J.47 Mantuo Castellano, 38; de Pilo, 38 manzanilla wine, 92, 93 marjoram, 118, 121, 123; sweet, 118, 121, 123 Marsala, 31, 59, 90, 91, 109, 111; composi- tion, 32 ; production of, 8 masterwort, 121, 123 Mataro, acreage, 9 maturity of grapes for harvesting, 35, 36, 37, 38, 39, 40, 74, 83, 92, 93, 95 meadowsweet, European, 120 mercaptans, 42 metal, contamination, 82, 89, 103, 105, 106, 112, 113, 126, 127, 147, 149; corrosion- resistant, 65, 67, 70, 71, 77, 112, 129, 133, 142; foil, 142; precipitation of, 148 microorganisms, see bacteria, Botrytis cinerea, and yeast milk, 130 Mission, 37, 38, 82; acreage, 9; for California Marsala, 110 mistelles, 24, 110 montiila, 92 montmorillonite, 132 Mourisco preto, 40 "mousiness," see spoilage Muscadelle, 36 muscat, raisining of, 83; second crop, 83; use for concentrate, 112; see Aleatico, Malvasia bianca, Muscadelle, Muscat Canelli, Muscat of Alexandria, Muscat Hamburg, Muscat St. Laurent, and Orange Muscat Muscat Canelli, 36, 81, 82, 119 Muscat of Alexandria, 26, 35, 83, 84; composi- tion of, 36 Muscat Hamburg, 36 Muscat St. Laurent, 36 muscatel, 9, 21, 24, 26, 28, 54, 55, 58, 59, 81-89, 95, 108, 112, 135, 137, 139, 140, 145; composition, 26, 28; composition of musts for, 35 ; flavor in white port, 24 ; in sherries, 104; in Spain, 95; in vermouth, 117, 123, 124; production of, 6, 8; red, 21, 73; special procedures for, 84—85; sulfur dioxide in fermentation of, 42 ; vari- eties for, 35, 36; volatile acid in, 14, 36, 44, 143 must, 41, 44, 77, 85, 102, 111 ; acetification of, 143; addition of concentrate to, 114; com- position of, 13, 14, 15, 16, 20, 33, 35-39, 40, 81, 95; deficiencies of, 40, 111; forti- fied, 24, 110, 136; free-run, 39, 41, 44, 81, 83, 84, 93, 113; lightly fermented, 42; lines, 65, 67, 71; pump, 65, 67; see also acidity, Balling, pH, and reduced must Mycoderma vini, 96, 98 ; see also yeast natural sweet wine, 16, 31; Hungarian Tokay, 111 nitrogen, 98, 99 ; influence on fermentation, 47 nomenclature, 21, 22, 92, 111; see also legal restrictions north coast, production in, 10 Norton, 106 nutmeg, 118, 119, 121, 123 oak, see containers, flavor, and wood oenin chloride, 20 Oloroso, 91, 92, 94, 96, 100, 101 optical rotation, 16, 44 Orange Muscat, 36 orange peel, bitter, 117, 118, 120, 123 orris, 117, 118, 121, 123 overfined wine, kaolin for, 132 oxidation, 4, 59, 61, 67, 68, 80, 90, 101, 102, 103, 109, 143; during sherry heating, 106, 107, 109; influence of low temperature on, Bul. 651] Commercial Production of Dessert Wines 185 69; inhibition by enzymes and proteins, 58; of aromatic principles in herbs, 119, 124; of tannin and coloring matter, 60, 126, 128, 129, 147, 148; over-, 26, 58, 84, 87, 98, 99, 103, 113; oxidation-reduction potential, 107, 147 ozonization, 61 Pagadebito, 40 palmas, 92, 94 (table) Palomino, 38, 93, 104; acreage, 9 parilla, 95 Pasteur, 98 pasteurization, 4, 15, 59, 61, 70, 77, 85, 88, 89, 105, 110, 127, 133, 142, 146; see also heating Pearson square, 52, 137-38 pectin, 85, 89 Pedro Ximenes, 38, 93, 95, 108 Perruno, 38 Petite Bouschet, 40 Petite Sirah, 39, 40 ; acreage, 9 pH, 13, 14, 47, 74, 104, 107; change during refrigeration, 128 ; influence on spoilage, 143, 146; of muscatel, 28; of musts, 35-38, 40 ; of port, 23 ; of sherry, 30 ; range for iron casse, 149 ; see also acidity phlobatannins, 148 phosphate, 89, 98, 99, 149 Pichia, 97 pipe, Madeira, 108, 109 ; port, 68, 80, 102, 134 pomace, 39, 44, 62, 65, 67, 77, 83; use for making distilling material, 56—57; still, 57 pomegranate, 118, 121, 123 port, 15, 16, 21, 54, 55, 59, 68, 73-81, 109, 112, 134, 142, 145; containers for, 80; composition of, 22, 23, 73 ; composition of musts for, 35 ; determination of color in, 162; fermentation period for, 41 ; for blend- ing, 31 ; precipitation of tartrates from, 128 ; production of, 6, 8 ; ruby, 22 ; special methods of making, 76, 77; sucrose in, 16; tawny, 21, 22, 24, 39, 138; use in making California Tokay, 111; varieties for, 38, 39, 40; "vintage," 22 Portugal, 16, 17, 22, 24, 27, 32, 39, 75, 77, 78, 79, 80, 108, 131 potassium metabisulfite, see sulfur dioxide potassium sulfate, in sherry, 93, 94 pressing, 44, 65, 66 (fig. 7), 67, 77, 78, 83, 84, 85, 93, 108, 110, 111 production, 4, 5 ; by districts, 10 Prohibition, 4, 6, 8, 9, 22, 23, 24, 25, 26, 27, 28, 30, 56, 79, 80, 81, 82, 83, 116, 134, 140 proteins, 20, 56, 58, 147; coagulation by heat, 110, see also colloids pumping over, 76, 79, 106, 110 puncheons, see containers punching down, 75, 76, 84 pure yeast culture, see yeast (use of pure cul- tures) quassia, 118, 121, 123 quema, 95 quinine, 33, 118, 125 racking, 15, 63, 80, 87, 95, 100, 108 raisining, 36, 38, 39, 76, 83, 93, 95, 110, 139 raisins, 45, 76, 83 : use for wine, 13 rancio, see flavor (rancio) raya, 92, 94, 95 red dessert wine, 21 ; composition of, 22, 23 ; composition of musts for, 35, 36, 38, 39, 40 ; directions for making, 73—81 ; see also muscatel (red), port, Tokay reduced musts, 16, 17, 31, 33, 40, 81, 91, 95, 114, 139 ; see also concentrate reducing sugar, see sugar refrigeration, 4, 61, 65, 69 (fig. 9), 60, 75, 80, 88, 110, 127, 147; of musts, 42, 143, 144; of vermouth, 124; to remove tartrates, 127 (table 31), 128, 129; see also stabilization resins, 101 ; in herbs for vermouth, 117, 118 rhubarb, 118, 121, 123; Chinese, 123; wine, 125 rosemary, 118, 121, 123 rot, 38, 74, 82 ; bunch, 39 Saccharomyces sp. 96, 97; cerevisiae var. ellipsoideus, 97; cheresiensis var. armensi- ensis, 96 ; see also yeast Sacramento Valley, production in, 10 "Saf," 131 saffron, 121, 123 sage, clammy, 120, 123 sake, 59, 144 Salvador, 39, 40 sanitation, in delivering grapes, 74, 82 ; in the winery, 65, 71, 72, 146; of fermenters, 76 San Joaquin Valley, 35, 74, 105 ; muscatel in, 8 Sanliicar de Barrameda, 92 sauterne, as dry vermouth base, 124 Sauternes, 3, 15 Sauvignon vert, 37; acreage, 9 savory, 121, 123 Scobicia declives, 73 seeds, 41, 57 settling, 58, 80, 131; of concentrate, 113; of yeast, 56 sherry, 21, 54, 55, 61, 68, 90-108, 137, 145, 149; acetal in, 19; aldehyde in, 19, 24, 29, 30; brown, 95; composition, 29, 30, 31; composition of musts for, 35 ; determination, of color in, 162; East India, 92; flavor, 90, 91, 92, 93, 97, 105; for blending, 31; golden, 92 ; muscat flavor in, 38 ; precipita- tion of tartrates from, 128; production of, 6, 8; room, 62 (fig. 3), 63, 64 (fig. 5), 72 (fig. 11), 104, 105; Spanish sherry yeast, 96, sucrose in, 17; tank for cooking, 71, 72 ; use in making California Tokay, 109 ; vari- eties for, 37, 38; see also California dry sherry, California sherry, California sweet sherry, and flavor (rancio) Sicily, 78, 109 ; see also Marsala skin, color in, 77; leaching of for flavor, 85; macerator for, 85, 86 (fig. 12) solera, 29, 96, 99, 100, 101, 125, 134, 135 South Africa, 29, 96, 101, 102, 144 southern California, 35, 81, 83; production in, 8, 10 Spain, 3 (footnote 5), 15, 21, 24, 29, 30, 38, 81, 91, 92, 93, 96, 97, 102, 108, 131, 132; Cordova, 92; Jerez de la Frontera, 90, 91, 93, 97, 98, 99, 109 ; Sanliicar de Barrameda, 92 Spanish earth, 124, 131, 132; see also ben- tonite specific gravity, 45, 46, 79, 113, 128, 157; table corresponding to Balling readings, 158 speedwell, 121, 123 spirits, 3 ; see also alcohol, brandy, and forti- fying brandy spoilage, 4, 13, 14, 15, 20, 36, 42, 44, 48, 57, 58, 72, 74, 80, 82, 84, 85, 95, 99, 103, 112, 125, 143-50; after fortification, 144-45; clarification to prevent, 126; "mousiness," 48, 144; nonbacterial, 147-50; of musts, 143-44; oxidasic, 150; see also bacteria and stabilization, 4, 111, 125, 126-29; see also filtration and fining State of California Department of Public Health, 14, 46 ; see also legal restrictions stemmers, see crushing sticking, see fermentation still, 57; capacity, 45; influence of sulfur dioxide on plates, 42; pomace, 57; room, 62, 63, 64, 67 succinic acid, 13 sucrose, 17, 34, 40, 46, 111, 119; in making caramel, 114—15; in native American spe- cies, 17; sweetness of, 16 sugar, 3, 15, 42, 75, 135, 137, 138, 139, 149; critical for film yeast, 98, 99, 103 ; determi- nation of, 158—60; in "aromatic" wines, 34; 186 University of California — Experiment Station in Angelica, 24, 25 ; in Madeira, 32 ; in Ma- laga, 30; in Marsala, 32, 110; in muscatel, 26, 28; in musts for dessert wines, 33, 35— 40, 45; in port, 22, 23, 73; in sherry, 14, 29, 30, 31, 92, 94; in vermouth, 34, 115, 117, 119, 124; in white port, 27; influence on expansion, 140; limits, 14, 21; not a factor in growth of lactic acid bacteria, 146 ; products from, 17, 18; relation to spoilage, 143 ; see also dextrose and levulose sulfur dioxide, 14, 41, 42, 57, 71, 73, 74, 75, 77, 84, 93, 98, 99, 102, 103, 104, 105, 108, 113, 124, 140, 144, 146, 147, 148, 149; determination of, 167, in muscatel, 28, 84; in sherry, 19, 30 ; influence on volatile acid, 41, 42; limit of error in determination of, 167 Sultanina, see Thompson Seedless sumps, 65, 66 sweet vermouth, see vermouth (sweet) sweet wine, see Angelica, muscatel, natural sweet wine, port, Tokay, white port Sylvaner, acreage, 9 table wines, 3, 35, 40, 41, 74, 90, 91, 128, 133, 145 ; consumption, 7, 8 ; definition, 3 ; pro- duction, 5, 6, 8, 61, 62 ; production by dis- tricts in California, 10 ; rate of aging, 58 tank cars, 72, 140 tannin, 20, 58, 59, 77, 81, 84, 90, 93, 101, 104, 128, 148; addition to wine, 108, 110, 111; determination of, 160; in "aromatic" wines, 34; in Angelica, 25; in fining, 131; in Madeira, 32 ; in Marsala, 32 ; in muscatel, 28; in port, 23; in sherry, 30; in Tokay, 32; in vermouth, 34, 117, 118; in white port, 27; limit of error in determination of, 167; oxidation of, 60, 126; precipitation of, 60, 126, 147, 148, 149 tartaric acid, 13, 19, 44, 59, 68, 111; for washing filter pads, 130 tartrates, 58, 59, 60, 69, 93, 113, 147; influ- ence of alcohol on, 12, 13; removal of, 127-28. tasting, 62, 92, 95, 100, 103, 107, 135, 136, 138—39 tawny port, 21, 22, 24, 39, 138 taxes, 6, 11, 167; on vermouth, 116 temperature, 60, 69, 75, 76, 77, 81, 84, 85, 102, 105, 107; for heating Madeira, 109; for heating sherry, 105, 106; influence of fortification on, 54 ; influence of size of containers on, 48; influence on clarity, 126 ; influence on fining, 131; influence on vol- ume, 139, 140; of fermentation, 41, 42, 57, 104; optimum for lactic acid bacteria, 146; see also heating and refrigeration thistle, blessed, 120, 123 Thompson Seedless, 9, 37, 38 thyme, 118, 121, 123 Tinta Cao, 39 Tinta Madeira, 39, 40 Tokay, 15, 110-11, 145; composition, 31, 32; production of, 6 Torulopsis dattila, 97 total acid, see acidity Touriga, 39 Trousseau, 38, 39, 40 ; for port, 21, 22 vacuum pan, 33, 112; types of, 112; use of, 113 Valdepefias, 39, 40 valerian, 118, 121 vanilla, 118, 121, 123, 124, 125 varieties of grapes, see grapes Verdelho, 37 vermouth, 3, 11, 33, 115-25 ; caramel sirup in, 118; clouding of, 124; composition, 34, 115; consumption, 12; dry (French) type, 33, 34, 115, 124, 125; herbs for, 117, 118, 119-22, 123; legal restrictions on, 116; method of preparation, 119, 122, 124; pos- sible medicinal properties of, 117; produc- tion, 12; room, 62 (fig. 3), 63, 116; sub- stances in, 117, 118, sweet (Italian), 33, 34, 115, 117, 119, 122-24 vin de liqueur, 3 vinho claro, 109 vinhos generosos, 109 vini dilusso, 3 vino, de color, 95, 108; de macetilla, 95; de pasto, 92 ; de roda, 109 ; generoso, 3 vintage dates, 68 ; see also port Vitis vinifera, 20, 73, 81 ; sugars in, 16, 17 volatile acid, see acetic acid whisky, use of sherry for coloring, 95 white dessert wine, 21, 24, 25 ; directions for making, 81—111; see also Angelica, Ma- deira, Malaga, Marsala, muscatel, sherry, Tokay, and white port white port, 55, 82-84, 85-89, 149; composi- tion of, 25, 26, 27; composition of musts for, 37 ; method of preparation, 88-89 ; varieties for, 37 wine, consumption, 7, 12 (table 5) ; economic status of industry, 4-11; export trade, 11; history, 4, 5, 22, 24, 26, 81, 116; produc- tion, 5 (table 1), 10, 12 (table 5); princi- ples of making dessert wines, 33—61 Wine Institute, 142, 168 wine thief, 138 winery, 62 (fig. 3), 63 (fig. 4) ; design, 62-64, 71; equipment for, 65-71, 129-30, 142; safety measures in, 72, 73; sanitation and maintenance of, 71—73 ; production of ver- mouth in, 116; tasting room in, 138 wood, extractives from, 20, 59, 60, 61, 68, 90, 101, 103, 104, 105, 108; see also con- tainers wormwood, 118, 121, 123, 125; Roman, 117, 121, 123 Xerez, sherry, Spain yarrow, 118, 121, 123 yeast, alcohol-tolerant, 47 ; apiculate, 47 ; autol- ysis, 44, 48, 58, 98; Brazilian Logos, 47; clouding due to, 43, 126, 146; esters formed by, 19, 61, 97; film, 29, 91, 96-103; respira- tory aerobic, 47 ; Saaz-type beer, 47 ; sau- ternes strain, 15 (footnote 19) ; sensitivity to alcohol, 47, 48, 143, 146; settling, 56, 87; strains, 41, 43, 47, 61, 75. 84, 96, 97; use of pure cultures, 43, 44, 65, 67, 74, 75, 84, 93, 104, 143 yeso, see gypsum zedoary, 121, 123 Zinfandel, 39, 40 ; acreage, 9 Zygosaccharomyces sp., 16, 18, 113, 146 18m-9,'41(5315)