GIFT OF 
 
 *? 
 
ABSCISSION OF FLOWERS AND FRUITS IN 
 
 THE SOLANACEAE, WITH SPECIAL 
 
 REFERENCE TO NICOTIAN A 
 
 A THESIS SUBMITTED IN PAETIAL SATISFACTION OF 
 THE EEQUIEEMENTS FOE THE DEGEEE OF 
 
 DOCTOE OF PHILOSOPHY 
 AT THE UNIVEESITY OF CALIFOENIA 
 
 BY 
 
 JOHN NORMAN KENDALL 
 
 MAY, 1917 
 
 
UNIVERSITY OF CALIFORNIA PUBLICATIONS 
 
 IN 
 
 BOTANY 
 
 Vol. .5, No. 12, pp. 347-428, 10 text figs., plates 49-53 March 6, 1918 
 
 ABSCISSION OF FLOWERS AND FRUITS IN 
 
 THE SOLANACEAE, WITH SPECIAL 
 
 REFERENCE TO NICOTIANA 
 
 BY 
 
 JOHN N. KENDALL 
 
 UNIVERSITY OF CALIFORNIA PRESS 
 BERKELEY 
 
UNIVERSITY OF CALIFORNIA PUBLICATIONS 
 
 Note. The University of California Publications are offered in exchange for the publi- 
 cation* of learned societies and institutions, universities and libraries. Complete lists of 
 all the publications of the University will be seat upon request. For sample copies, lists of 
 publications and other information, address the Manager of the University Press, Berkeley, 
 California, U. 8. A. All matter sent in exchange should be addressed to The Exchange 
 Department, University Library, Berkeley, California, U. 8. A. 
 
 BOTANY. W. A. Setcheli, Editor. Price per volume. $8.50. Volumes I (pp. 418), U (p* 
 
 360), III (pp. 400), IV (pp. 379), completed. Vols. V, VI and Vn in 
 progress. 
 
 Cited as Univ. Calif. PubL Bot. 
 
 VoL 1. LA Botanical Survey of San Jacinto Mountain, by Harvey Monroe 
 
 HalL Pp. 1-140; plates 1-14. June, 1902 $1.00 
 
 2. Two new Ascomycetous Fungi Parasitic on Marine Algae, by Minnie 
 
 Eeed. Pp. 141-164; plates 15-16. November, 1902 .2S 
 
 3. Algae of Northwestern America, by William Albert SetcheU and Na- 
 
 thaniel Lyon Gardner. Pp. 165-418; plates 17-27. March, 1903 2.2S 
 
 Vol. 2. LA Review of California Poleinoniaceae, by Jessie Milllken. Pp. 1- 
 
 71; plates 1-11. May, 1904 _ .75 
 
 8. Contributions to Cytological Technique, by W. J. V. Osterhout. Pp. 
 
 73-90; 5 text-figures. June, 1904 .25 
 
 8. Limu, by William Albert SetcheU. Pp. 91-113. April, 1905 _ .25 
 
 4. Pest-Embryonal Stages of the Laminariaceao, by William Albert 
 
 SetcheU. Pp. 115-138; plates 13-14. April, 1905 . .25 
 
 5. Regeneration among Kelps, by William Albert SetchelL Pp. 139-168; 
 
 plates 15-17. July, 1905 .. .30 
 
 6. A New Genus of Ascomycetous Fungi, by Nathaniel Lyon Gardner. 
 
 Pp. 169-180; plate 18. July, 1905 _ .15 
 
 7. Teratology in the Flowers of some California^ Willows, by William 
 
 Warner Mott. Pp. 181-226; plates 16-20. December, 1905 .60 
 
 8. 9, 10. 11. (In one cover.) The Resistance of Certain Marine Algae to 
 
 Changes in Osmotic Pressure and Temperaturs. The ROle of Os- 
 motic Pressure in Marine Plants. On the Importance of Physiolog- 
 ically Balanced Solutions for Plants. The Antitoxic Action of 
 Potassium on Magnesium. By W. J. V. Osterhout. Pp, 227-236. 
 
 March, 1906 _ _ _ . 3& 
 
 12. Cytological Studies in Cyanophycea*, by -Nathaniel Lyon Gardner. 
 
 Pp. 237-296; platea 21-26. November, 1906 _. LOO 
 
 15. On a Small Collection of Mosses from Alaska, by J. Cardot and T. 
 
 Theriot. Pp. 297-308; plates 27-28. December, 1906 .10 
 
 14. Some Uureported Alaskan Sphagna, together with a Sun- jury of the 
 Cryptogamic Work of the University of California Botanical Ex- 
 pedition to Alaska in 1899, by William Albert SetcaelL Pp. SOS- 
 SIS. September, 1907 _ .06 
 
 16. On Nutrient and Balanced Solutions, by W. 3, V. Osterhout. Pp. 317- 
 
 318. October, 1907 ^ .06 
 
 16. A Synopsis of the North American Godeti&s, by Willis Linn Jepsoa. 
 
 Pp. 319-554; plate 29. December, 1907 ..... AO 
 
 Index, pp. 355-300. 
 
 . 1907-1009. 
 
 L Compositae of Southern California, by Harvey Monroe Hall. Pp. 1- 
 
 302; plates 1-3, with a map. December, 1907 9,00 
 
 2, The Origin, Structure, and Function of the Polar Caps in Sm&acina 
 
 amplexicaulis Nutk, by H. D. Densmore. Pp. 803-330; plates 4-8. 
 
 December, 1908 .89 
 
 8, 4. (In one cover.) The Value of Sodium to Plants by Reason of It 
 
 Protective Action. On the Effects of Certain Poisonous Gases on 
 
 Plants. By W. J. V. Ostsrhout. Pp. 331-340. June, 1908 . _ .10 
 
 6. Contributions to the BInowledge of the California Species of Cmsta- 
 
 CCOBS Corallines. L by Maurice Barstow Nlchola. Pp. 341-348; 
 
 plate 9. December, 1908 .. .10 
 
UNIVERSITY OF CALIFORNIA PUBLICATIONS 
 
 IN 
 
 BOTANY 
 
 Vol. 5, No. 12, pp. 347-428, 10 text figs., plates 49-53 March 6, 1918 
 
 ABSCISSION OP FLOWERS AND FRUITS IN 
 
 THE SOLANACEAE, WITH SPECIAL 
 
 REFERENCE TO NICOTIAN A 
 
 BY 
 
 JOHN N. KENDALL 
 
 CONTENTS 
 
 PAGE 
 
 I. Introduction 348 
 
 II. Summary of the literature 350 
 
 III. Technique 361 
 
 IV. Histology and cytology of the pedicel 363 
 
 1. Histological and cytological condition of the mature pedicel 363 
 
 2. Development of the separation zone in Nicotiana and Lycoper- 
 sicum 367 
 
 3. Increase in size and development of mechanical tissue in the 
 pedicel of Nicotiana and Lycopersicum 369 
 
 V. The process of abscission 371 
 
 1. General description of the process in several genera 371 
 
 2. Method of cell separation 376 
 
 VI. Abscission of the style and corolla 383 
 
 VII. Time of abscission 385 
 
 1. Keaction time 385 
 
 2. Abscission time 396 
 
 VIII. Experimental induction of abscission 397 
 
 1. Induction by illuminating gas 397 
 
 2. Action of acids on the separation layer of Nicotiana 404 
 
 3. Induction by mechanical injury 406 
 
 4. The ability of certain species to throw off pedicels from which 
 
 all the floral organs have been removed, as related to the induc- 
 tion of abscission by mechanical injury 410 
 
 IX. Summary 411 
 
 X. Conclusion 415 
 
 XI. Literature cited 418 
 
 XII. Plates .. 420 
 
 374794 
 
- 
 
 ? *; : % * / 
 . 
 
 348 University of California Publications in Botany [VOL. 5 
 
 INTRODUCTION 
 
 Although it is a matter of common observation that many plants 
 are capable of detaching portions of the body, the underlying cause 
 and the actual mechanism which bring about such separation are 
 only slightly understood. The process has often been described as 
 one of self-pruning by which the plant rids itself of useless portions 
 of its body. Since abscission is sometimes confused with exfoliation, 
 it seems desirable here to distinguish definitely between these two 
 phenomena. It can be said that, in general, exfoliation is preceded 
 by drying and death of the part to be cast off and that actual separa- 
 tion of the organ is accomplished by a mechanical break through dry, 
 dead tissues. Abscission, on the other hand, is usually not preceded 
 by drying and death of the organ concerned and its detachment is 
 accomplished by a separation along the plane of the middle lamellae of 
 active living cells. 
 
 Abscission may be either axial or lateral. Axial abscission includes 
 the abscission of portions of stems, shoots, entire flowers or fruits. 
 Lateral abscission includes the abscission of leaves, petioles, sepals, 
 petals or styles. Considerable attention has been given by investi- 
 gators to the abscission of flowers because of the theoretical detriment 
 to crops caused by the fall of the flower before the fruit is formed. 
 
 The cause of leaf-fall in deciduous species is connected with peri- 
 odic changes in the physiological condition brought about by changes 
 in the environment. In the case of some herbaceous plants and occa- 
 sionally in trees, sudden changes in environmental conditions result- 
 ing in a loss of physiological equilibrium often cause the throwing 
 off of leaves, flowers or even small shoots. In certain species, any- 
 thing which tends to loss or completion of function within or peculiar 
 to an organ causes the organ to be thrown off. Thus, staminate flow- 
 ers are commonly thrown off soon after anthesis and pistilate flowers 
 generally fall when fertilization is prevented. Similarly, certain 
 species e.g., Impatiens Sultani and Mirabilis Jalapa throw off por- 
 tions of their stems which have been rendered useless as a part of the 
 conducting system because of injury or removal of distal buds or 
 leaves. 
 
 The following definitions of terms, which will be used throughout 
 this paper, are made necessary because of a notable lack of uniformity 
 in their usage by various investigators who have dealt with abscission. 
 
1918] Kendall: Abscissi-on of Flowers and Fruits in Solanaceae 349 
 
 1. Abscission is the detaching of an organ by the separation of 
 actively living cells at or near its base. 
 
 2. The separation layer (Mohl's Trennungschichte) is the layer of 
 cells the components of which will separate from one another at 
 abscission. 
 
 3. The separation cells or absciss cells are the cells that make up 
 the separation layer. 
 
 4. The separation zone is the general region through which abscis- 
 sion takes place and usually is largely proximal to the separation layer. 
 
 A preliminary account of abscission in F x species hybrids of Nico- 
 tiana has already appeared (Goodspeed and Kendall, 1916). The 
 present study represents an amplification of this investigation and its 
 extension to other species of the Solanaceae. It is particularly con- 
 cerned with the following: (1) the position of the separation layer; 
 (2) the origin of the separation layer; (3) the cytology of the separa- 
 tion layer; (4) the process of abscission, including (a) a description 
 of the appearance of the separation layer in consecutive stages of the 
 process and (b) the method of cell separation; (5) the time occupied 
 by abscission, including (a) the time between the application of the 
 stimulus and fall (reaction period) and (b) the time involved in the 
 actual process of cell separation (abscission period) ; (6) experimental 
 induction of abscission. 
 
 Although the investigation reported here is largely a morpholog 
 ical one, the results of the experiments on the method of cell separa- 
 tion, the time of abscission and the induction of abscission seem to 
 have a distinct physiological significance as well. 
 
350 University of California Publications in Botany [VOL. 5 
 
 SUMMARY OF THE LITERATURE 
 
 Since the literature on abscission is rather voluminous, it seems 
 best to present the following discussion under several different head- 
 ings corresponding, to a certain extent, with the six main topics of 
 interest mentioned in the introduction. The summary below is largely 
 confined to the literature on axial abscission, although that on lateral 
 abscission is considered in so far as it has a direct bearing on the most 
 important aspects of the abscission problem. 
 
 1. HISTOLOGY OF THE PEDICEL 
 
 a. POSITION OF THE SEPAEATION LAYER 
 
 Hoehnel (3880), discussing the fall of catkins in Populus and 
 Salix, locates the separation layer at the base of the catkin. The gen- 
 eral region at the base of the catkin, in the distal part of which the 
 separation layer is located, he calls the "separation zone." In Salix, 
 actual separation occurs in the separation layer, but in Populus it 
 occurs in the parenchyma entirely outside the separation layer. 
 According to Balls (1911), the separation layer in the cotton flower 
 is located at the base of the pedicel. The layer is located by Hannig 
 (1913) at the base of the pedicel in Nicotiana Tabacum, N. rustica, 
 N. accuminata, N. sylvestris, Datura, and Atropa, and at the tip of 
 the pedicel in Nicotiana Langsdorffii, Salvia Aloe, Cuphea, and Gasteria. 
 He finds it occurring at the middle of the pedicel in Impatiens Sultani, 
 Solanum tuberosum, Lycopersicum, Asparagus, and Begonia. Gort- 
 ner and Harris (1914) and Lloyd (1914&), working on the abscission 
 of internodes as the result of injury in Impatiens Sultani, locate the 
 separation layer at the first node below the injury and just above the 
 axillary bud. Occasionally, according to the latter investigators, ab- 
 scission may occur at the second or third node below the injury and 
 in these cases the buds at the first or second nodes seem to be abortive. 
 
 The separation layer, according to Hannig (1913), may occur at 
 the base of the complete inflorescence in Impatiens and Oxybaphus. 
 According to Lloyd (1914a), the separation layer occurs at the base 
 of the pedicel in cotton and at the base of the ripened ovary in grape 
 "shelling." In the abscission of internodes and tendrils in Vitis and 
 Ampelopsis, Lloyd (19140) locates the layer near but not exactly at 
 the base of the internode. A peculiar case illustrating the result of 
 displacement of the stem on the location of the separation layer is 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaveae 351 
 
 discussed by Lloyd (1914a) for Ampelopsis and Gossypium. In the 
 latter, abscission, in the abnormal case, occurred down the internode 
 at the base of the pedicel. This is explained as the result of a dis- 
 placement during growth by which part of the pedicel becomes united 
 to the stem. 
 
 Occasionally, grooves or swellings are noticed at the base of the 
 organ being abscissed where they correspond more or less" exactly to 
 the general position of the separation layer. Examples are given by 
 Hannig (1913) for Lycopersicum and Solanum tuberosum and by 
 Balls (1911) for Gossypium. Abscission may occasionally occur, 
 according to Lloyd (1914a), above a small bract. According to these 
 latter investigators, there is more often no external indication of the 
 layer. Frequently, grooves bear no relation to the layer because in 
 many cases of this kind (Hannig, 1913, for Brunfelsia) separation 
 occurs a short distance distal to the groove. 
 
 From the above brief summary it is evident that in the case of 
 axial abscission the separation layer is located at or near the base of 
 an internode. Apparent exceptions are reported by Hannig (1913) 
 in which it is seemingly located at the middle of an internode. It 
 seems probable that a more critical re-examination might reveal the 
 fact that even these exceptions accord with the general rule. In these 
 cases, for example, the pedicel of the flowers in question might be 
 composed of two internodes. 
 
 6. ORIGIN OF THE SEPAEATION LAYEE 
 
 Kubart (1906) states that the occurrence of the separation layer 
 in all tyes of abscission may be explained in one of the three following 
 ways: (a) the separation layer is preformed and represents simply 
 a portion of the primary meristem which has remained in its original 
 active state; (6) it represents a secondary meristem; (c) the primary 
 meristem may function directly as a separation layer. The differ- 
 ence between a and c is only a difference in time, c being added to 
 explain the origin of the separation layer in abscission of very young, 
 embryonic tissues. In a, the separation layer is present at the base 
 of the organ from the start of its development, but in & this layer has 
 to be formed by a secondary meristem before abscission can occur. In 
 a, cell divisions are not normally found preceding abscission, but in & 
 and c they are. Mohl (1860), working on the fall of the flower in 
 Aesculus, Pavia, Lagenaria, Cucumis, and Ricinus, states that the 
 separation layer in these forms is of type 6. Throughout his entire 
 
352 University of California Publications in Botany [ V L. 5 
 
 work Mohl gives the general impression that it is necessary for a sep- 
 aration layer to be formed from a secondary meristem before abscis- 
 sion can occur. Wiesner (1871), working on leaf -fall in general, 
 observes that the separation layer is not generally of type b, as Mohl 
 believes, but more often of type a. According to Becquerel (1907), 
 the separation layer is formed in the pedicel of Nicotiana from a sec- 
 ondary meristem (type b). In the cotton flower Balls (1911) finds 
 that the separation layer is of type Z>, but according to Lloyd (1914a 
 and 1916&) there is doubt as to this conclusion, since in the case of 
 very young cotton flowers in which abscission occurs very suddenly, 
 he finds only rarely that cell divisions do not precede abscission. 
 Hannig (1913), for flower-vfall in general, states that a separation 
 layer of type a is always present but in certain species a secondary 
 layer of type b may also be formed, through which separation may or 
 may not occur. Hannig, differing from Becquerel ( 1907 ) , points out 
 that the separation layer in Nicotiana is of type a. Lloyd (1914a) 
 and Loewi (1907) indicate that in general a layer of cells through 
 which abscission is possible is more often of type a than of type b. 
 They believe, however, that the separation layer is not a definite 
 morphological structure but represents merely a physiological con- 
 dition. 
 
 c. CYTOLOGY OF THE SEPAKATION LAYEE 
 
 Mohl (1860) describes the separation cells in the flower stalk as 
 young, active, small cells which generally contain no starch. He also 
 states that in most cases cell divisions are characteristic of the sep- 
 aration layer, i.e., that the separation layer is meristematic. Hoehnel 
 (1880) finds that cell divisions are characteristic of the proximal por- 
 tion of the separation zone in Salix and Populus but in the distal 
 portion, where the separation layer is developed, these divisions are 
 not so numerous. In some cases he finds separation taking place in 
 the parenchyma, entirely outside the "zone" where there had been 
 no cell divisions. The separation cells in Nicotiana are described by 
 Becquerel (1907) as small, practically undifferentiated cells with 
 large nuclei. In Begonia, Fuschia, Mirabilis, and Impatiens Hannig 
 (1913) describes the tissue as secondary meristem (type b) with the 
 cells rectangular in shape and arranged in more or less definite rows. 
 In contrast to the above observations, he describes the cells as small, 
 irregularly arranged and spherical in Salvia, Solanum nigrum, and 
 Nicotiana Tabacum. In Solanum nigrum the separation layer consists 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 353 
 
 of two or three tiers of cells but in N. Tabacum the layer is made up 
 of ten to fifteen tiers. 
 
 Hannig (1913), by means of various microchemical tests, can 
 detect no chemical difference between the cell walls of the separation 
 layer and those of the cells on either side. Lloyd (1914a), however, 
 claims that the cell walls of the separation cells break down more 
 quickly when treated with caustic potash than do the walls of normal 
 cells. Starch grains are frequently noted by Hannig and Lloyd 
 (1916a) as occurring in the separation cells, especially in the abscis- 
 sion of internodes by Mirabilis Jalapa. 
 
 An examination of the literature thus makes it evident that there 
 has been a great difference noted in the various species in regard to 
 the character of the separation cells. The one characteristic of these 
 cells, however, to which there is no exception is that they are in an 
 actively living condition. 
 
 2. THE PROCESS OF ABSCISSION 
 
 a. METHODS OF ABSCISSION 
 
 It has been found that in practically all cases of abscission the 
 detaching of the organ is brought about by the separation of cells 
 along the plane of the middle lamella. It is the method noted by 
 Mohl (1860), Wiesner (1871), and Kubart (1906), who call it a pro- 
 cess of maceration. Correns (1899) calls it a process of ' ' schizolysis. " 
 Correns, however, in the same work describes a new and different 
 method of abscission (rhexolysis) which he finds in mosses. In this 
 latter method, separation is accomplished by a seemingly passive 
 break of tissues irrespective of the position of cell walls. This may 
 be the case in the style of cotton (cf. Lloyd, 1914&). This same 
 method has been reported by Tison (1900) in the leaf of Aristolochia 
 Sipho, although the evidence has been called in question by Lloyd 
 and Loewi (1907). Still another type of abscission has been described 
 by Hannig (1913) as a result of experiments on Mirabilis and Oxy- 
 baphus. In these plants he finds separation being brought about by 
 a disorganization and dissolving away of a complete tissue. Lloyd 
 (1916a), on the other hand, states that separation in these species is 
 accomplished by cell separation and is thus true schizolysis. Hannig 
 was doubtless confused in this case by the cell elongations which 
 Lloyd observes and by which the membranes surrounding the proto- 
 plasts are drawn out exceedingly thin. Loewi (1907), working on 
 
354 University of California Publications in Botany [VOL. 5 
 
 several genera, including Cinnamomum and Euonymus, notes and 
 figures cell elongations similar to those figured by Lloyd (1916a). 
 These cell elongations he finds so frequent and conspicuous that he 
 proposes a distinct type of abscission, calling it a " Schlauchzell 
 raechanismus. ' ' 
 
 Loewi, on the basis of his studies, seeks to classify the methods of 
 cell separation in abscission under six different headings, which per- 
 haps would be more appropriately presented under the next subject 
 of consideration (the methods of cell separation) ; but since the author 
 gave them as distinct methods of abscission they will be considered 
 here. They are: (1) "round cell" mechanism; (2) dissolving of the 
 middle lamella; (3) maceration; (4) turgescence; (5) cell elonga- 
 tions; (6) "hard cell" mechanism. They are to be considered merely 
 as factors which, singly or in combinations, may enter in as a part of 
 the normal process of cell separation. Loewi also claims that by con- 
 trolling the temperature, humidity, and various other factors sur- 
 rounding the plant he can influence it to such an extent as to change 
 its method of cell separation. 
 
 ft. METHOD OF CELL SEPARATION 
 
 It has been held by various investigators that the cell separation, 
 almost universally connected with abscission, can be caused either by 
 (a) chemical alteration and dissolving of the middle lamella or by 
 (&) increase in cell turgor. This whole matter has received consider- 
 able attention, although very little direct evidence has been obtained. 
 Wiesner (1871 and 1905) states that cell separation is caused by the 
 dissolution of the middle lamella and by increased turgor. Kubart 
 (1906) and Loewi (1907) agree entirely with Wiesner on this point. 
 Strasburger (1913), Tison (1900), Lee (1911), Hannig (1913), and 
 Lloyd (1916# and b) believe that cell separation is accomplished by 
 the dissolution of the middle lamella. Practically all investigators 
 have noticed the turgid appearance of the cells after separation, 
 although this of course does not constitute evidence that the separa- 
 tion is due to increased turgor. Fitting (1911) claims that the sep- 
 aration is accomplished, at least in some cases, solely by an increased 
 turgor of the separation cells. He bases his claim on the fact that 
 abscission is very often too rapid to allow time for the dissolution of 
 the middle lamella. He also mentions the fact that the separation 
 cells are very often small, spherical cells, the type of cell which would 
 respond most readily by an increase in cell turgor. On account of its 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanac&ae 355 
 
 rapidity and regularity of reaction, Fitting claims that abscission is 
 a semi-tropistic phenomenon and suggests the term "Chorismus" to 
 designate this type of reaction. 
 
 It has been observed by Hannig and Fitting that the presence of 
 various narcotic vapors in the atmosphere around certain species of 
 plants causes their flowers or merely the petals to be thrown off. 
 Various aspects of this general problem of the reaction ofplant tissues 
 to such agencies have been investigated. It has been determined by 
 various plant physiologists that the presence of narcotic vapors, such 
 as illuminating or acetylene gas, in the air around certain plant tissues 
 causes the proportion of soluble carbohydrates within their cells to 
 increase. This increase in the amount of soluble carbohydrates would 
 indicate an increase in cell turgor. The question at once arises, 
 whether or not this increase in turgor can effect complete separation 
 or maceration of cells without the occurrence of chemical alteration in 
 the walls. Kichter (1908) resting his case on experimental evidence, 
 throws some light on this problem. Various kinds of plant tissues 
 which he subjected to acetylene vapors broke in pieces because of the 
 maceration and collapse of the living cells within. He finds that in 
 the case of the cells of tissues which are commonly rich in starch 
 inclusions, such as the fruit of the snowberry and the potato tuber, 
 the maceration is most complete. In the potato, for example, 3 to 
 5 mm. of material on the surface become completely macerated after 
 being subjected to acetylene gas. According to Bichter and Grafe 
 (1911), the proportion of sugar in starchy seedlings subjected to 
 acetylene gas is larger than in seedlings grown under normal condi- 
 tions. In seedlings from oily seeds, however, the amount of sugar is 
 decreased and the proportion of glycerine and fatty acids increased. 
 The conclusion is therefore drawn that the subjection of plant tissues 
 to narcotic vapors favors the hydrolysing process in the cells involved. 
 The work of these two investigators goes to show that narcotic vapors 
 may cause abscission by acting in either of the most important meth- 
 ods suggested as responsible for cell separation ; they may increase cell 
 turgor on the one hand or favor the hydrolysis of the middle lamella 
 on the other. 
 
 Lloyd (1916a) presents evidence of chemical change in the cell 
 walls of the separation layer before abscission. These cell walls stain 
 in the usual manner with iodine, giving a light brownish color, but 
 as abscission commences, they give a faint blue color when stained 
 with iodine and washed out with water. Shortly before cell separa- 
 
356 University of California Publications in Botany [VOL. 5 
 
 tion commences, Bisrnark brown and Ruthenium red fail to stain the 
 primary and secondary cellulose membranes of the separation cells, 
 although, when abscission does not occur, the entire cell wall is stained 
 in the normal manner. The cells when separating seem, furthermore, 
 to be surrounded only by the thin tertiary membranes. Lloyd, in his 
 work, figures cells in the process of separation which show the disso- 
 lution of the primary and secondary membranes of the cell wall. 
 
 Various interpretations are given to the repeatedly observed 
 occurrence of cell divisions preceding and accompanying abscission 
 Mohl (1860) expresses the opinion that cell divisions are generally 
 necessary before abscission can occur. Investigators since his time 
 have disproved the universal occurrence of cell divisions because they 
 find more and more cases where no cell divisions occur. Lloyd 
 (1914a) maintains that cell divisions are not of necessity correlated 
 with abscission but are merely evidences of renewed growth and 
 wound responses. As evidence he states that cell divisions are some- 
 times absent and sometimes present in the same species. He cites 
 (19165) the cotton plant as a typical example in which cell divisions 
 are present in the abscission of older flowers in which the reaction to 
 stimulus is slow. In young flowers and flower buds abscission may 
 proceed without cell division. He further notes (19140) that cell 
 divisions sometimes precede and at other times follow abscission in a 
 given species. 
 
 c. AGENCIES ACTIVE IN BEINGING ABOUT THE DISSOLUTION 
 OP THE MIDDLE LAMELLA 
 
 Very few theories have been proposed to account for the dissolu- 
 tion of the middle lamella and practically no evidence of any kind 
 has been submitted. "Wiesner (1905) claims that in leaf -fall an 
 organic acid, produced as a result of lessening of cell activity and 
 stagnation of cell contents, acts on the middle lamella. His evidence 
 for this statement has to do with obtaining acid reactions with litmus 
 from cells at the base of the petiole during abscission. Kubart (1906) 
 also obtains acid reactions at the base of the corolla in Nicotiana dur- 
 ing abscission and, although agreeing with Wiesner that an organic 
 acid probably causes the dissolution of the middle lamella, he also 
 admits the possibility that an enzyme plays a part in the process. 
 Lloyd (1916&) makes the statement that the dissolution of the middle 
 lamella is a process of hydrolysis and although making no definite 
 statement on the subject appears to take it for granted that an 
 
1918 ] Kendall: Abscission of Flowers and Fruits in Solanaceae 357 
 
 enzyme of some kind is the active factor. Indeed, since all hydrolys- 
 ing processes of living cells are now supposed to be due* to the action 
 of enzymes, there is no reason to suppose that the hydrolysis of the 
 middle lamella does not conform to the general rule. For it is known 
 that an enzyme, pectosinase, is capable of breaking down the pectose 
 of which the middle lamella is composed. However, until more is 
 known concerning the nature of this particular enzyme ~it remains 
 impossible to get more definite evidence on this phase of the problem. 
 
 3. ABSCISSION OF THE COROLLA 
 
 Reiche (1885) gives an account of the fall of the corolla in a 
 large number of species belonging to about forty-five families of the 
 monocotyledons and dicotyledons. He finds that the corolla may be 
 thrown off in one of three different ways: (1) by the activity of a 
 small-celled separation layer; (2) through decay; (3) through in- 
 crease in size of the ovary, thus tearing off the tissue involved at 
 the base of the corolla. In many cases of true abscission case 1 
 above Reiche finds that the separation layer is preformed and ready 
 to function at any moment. This represents a contradiction of 
 Mohl's observations, according to which the fall of the corolla is 
 usually due to the action of a separation layer formed shortly before 
 fall. According to Reiche, the separation layer is very seldom morpho- 
 logically differentiated from the neighboring tissue, but in a few cases 
 he describes the separation layer as consisting of a layer of cells 
 smaller than the neighboring cells on either side. 
 
 Kubart (1906), in his account of abscission of the corolla in sev- 
 eral different species, describes and figures the process which takes 
 place in Nicotiana. The separation layer in this genus he finds to be 
 in no way morphologically differentiated, of indefinite shape, and 
 located about 1 mm. above the base of the corolla tube. In this gen- 
 eral region a large number of cells separate from one another, all the 
 cells in cross-section taking part except the epidermal cells and the 
 tracheae. Fitting (1911), in his work on the shedding of petals, de- 
 scribes the process of abscission in several genera, paying particular 
 attention to Erodium, Geranium, Linum, Helianthemum, Perlagonium, 
 and Verbascum. Separation in these cases takes place through a 
 region of small, spherical cells rich in protoplasm. The separation 
 layer is not sharply differentiated as compared with the tissues on 
 either side but is located in a restricted region at the base of the petal. 
 
358 University of California Publications in Botany [VOL. 5 
 
 He finds no cell divisions preceding or accompanying abscission. The 
 process in premature abscission he finds differing in no way from 
 that in normal abscission after fertilization. These conditions, he 
 states, correspond more or less to those which he finds in the pedicel 
 during flower-fall. 
 
 4. TIME OF ABSCISSION 
 
 The time elapsing between anthesis and flower-fall in partially 
 sterile F 1 species hybrids of Nicotiana and between emasculation at 
 anthesis and fall in the case of their corresponding parents is dis- 
 cussed in a previous paper (Goodspeed and Kendall, 1916). It was 
 there stated that the average time is about nineteen days in F l H154, 
 seven in F x H179, five in N. Tabacum var. macrophylla, and thirteen 
 in N. sylvestris. When we turn to the question of the reaction time 
 in premature abscission occurring before the normal time as the result 
 of sudden changes in external environmental conditions, we find that 
 this subject has received only slight attention. According to Lloyd 
 (1914a), the cotton "square" falls in one to twenty-two days after 
 the weevil lays its eggs, the average time being eight days. In one 
 experiment in which the ovary was cut transversely, Lloyd was able 
 to cause one hundred per cent of the young bolls to fall in forty-eight 
 hours and ninety per cent in twenty-four hours. Larger bolls take a 
 longer time to respond to injury than do smaller ones, as a result of 
 the development of the pedicel to a condition in which abscission 
 meets greater resistance. Cotton * ' squares, ' ' he finds, take a longer time 
 to respond than young bolls, the former shedding thirty-five to sixty 
 per cent in thirty-six hours and the latter forty to seventy per cent in 
 forty-eight hours. On the other hand, he obtains no evidence (19166) 
 that the reaction times are any shorter in small buds than in larger 
 ones. The reaction times in cases where the injury is performed in the 
 evening seem to be shorter by about twelve hours than in cases where 
 the injury is performed in the morning. This difference he ascribes 
 to the increase in turgidity which takes place during the night and 
 which serves to hasten the reaction. Very severe injuries to the ovary, 
 he finds, cause fall of young bolls quicker than less severe injuries. 
 Injuries which are less severe than those mentioned above and per- 
 formed so as to imitate the injury inflicted on the ovary by insect 
 larvae caused shedding in three to six days, with most of the fall 
 occurring on the fifth day. Summing up his entire results, Lloyd 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 359 
 
 (1916&) states that under field conditions the responses to all kinds 
 of stimuli conducive to abscission become evident within ten days, 
 with the maximum frequency below six days. 
 
 The actual time involved in the process of abscission (abscission 
 time) has received even less attention than the problems discussed 
 above. Fitting (1911) states that abscission time may occasionally 
 be very short, forty-five seconds to five minutes in the petals of Ver- 
 bascum and thirty seconds to six minutes in Geranium. Lloyd (1914a 
 and 19166) finds abscission after injury of the small cotton-boll taking 
 place within four hours, the length of time depending somewhat on 
 the age of the boll. In a previous paper (Goodspeed and Kendall, 
 1916) a general estimate of the abscission time was given and it was 
 stated that normal abscission due to lack of fertilization takes place 
 in Nicotiana hybrids in four to eight hours and premature abscission 
 in one to four hours. 
 
 5. EXPERIMENTAL INDUCTION OF ABSCISSION 
 
 According to Hannig and Loewi, abscission may be induced in 
 two different ways. First by abnormal external conditions ("spon- 
 taneous" or premature abscission) and second by normal internal 
 conditions at the normal time ("automatic" or normal abscission). 
 We shall consider in the following summary of the literature only 
 two aspects of induction of the first type. 
 
 a, INDUCTION BY NAECOTIC VAPOES 
 
 Hannig (1913) reports a comparative study of the behavior of 
 cut sprigs of different species of plants when subjected to laboratory 
 air and to illuminating gas. He notes the fact that under either of 
 the above conditions all the flowers and occasionally a few small 
 shoots are abscissed. He finds, however, that not all the species in a 
 given family behave similarly in response to these conditions. We 
 are particularly interested in the Solanaceae and we may note that 
 this family contained more species that detached their flowers in 
 illuminating gas than any other of the families investigated by Han- 
 nig. According to Fitting (1911), narcotic vapors such as tobacco 
 smoke, carbon dioxide, ether, chloroform or illuminating gas fre- 
 quently cause premature abscission of the corolla. He notices, how- 
 ever, that ammonia or turpentine vapors fail to cause abscission. 
 Brown and Escomb (1902) make the statement that Nicotiana, Cu- 
 curbita, and Fuchsia shed flowers and buds in air containing only 
 0.114 per cent carbon dioxide. 
 
360 University of California Publications in Botany [ V <> L - 
 
 fc. INDUCTION BY MECHANICAL INJUEY 
 
 Becquerel (1907), in a brief paper on the effect of wounding 
 flowers of Nicotiana, notes that even after fifteen days flowers without 
 sepals, anthers, or stigmas do not fall. After the same length of time, 
 flowers without corollas or flowers in which the corolla or stamens are 
 only half removed, have fallen. He points out that this result is more 
 conspicuous in young flowers but did not investigate this point suffi- 
 ciently to arrive at any definite conclusions. According to Hannig, 
 removal of various organs of flowers frequently causes abscission but 
 wounding of the pedicel does not. He concludes, therefore, that in- 
 jury itself does not cause abscission but only acts indirectly by inter- 
 fering with important physiological processes in the treated tissues. 
 
 According to Lloyd (1914a), shedding of very young cotton-bolls 
 can be induced by removal of the styles before pollination, but fall in 
 this case can be assigned, as Fitting has shown, to lack of fertilization. 
 It appears that in the cotton flower (Lloyd, 1916&) there is an inhibi- 
 tion period which starts with the opening of the corolla and during 
 which premature abscission as the result of sudden stimuli very sel- 
 dom occurs. Also, cotton-bolls larger than 30 mm. in diameter are 
 very seldom shed under any conditions. Other results obtained by 
 Lloyd on the effect of injury on the abscission of cotton flowers are 
 discussed above under "Time of Abscission" (page 357). Lloyd 
 (1914&) also notes the effect of injury on abscission of internodes in 
 Impatiens Sultani. Plants of this species, when a cut is made across 
 the stem, cast off the remainder of the severed internode. He gives 
 results of experiments on the effect of different types of injury, noting 
 that some severe injuries do not cause abscission. Gortner and Harris 
 (1914) have obtained similar results with the same species. They 
 find that when the cut is made across the internode, very close to 
 the separation layer, abscission usually occurs, but occasionally it does 
 not. They state, as does Lloyd, that the shape and location of the 
 separation layer may vary slightly according to the type of injury. 
 
 c. THE DIEECT OE INDIEECT ACTION OF THE EXTEENAL 
 STIMULUS 
 
 In all the above investigations the question naturally arises, 
 whether the narcotic vapors and injuries or any stimulus conducive 
 to abscission act indirectly through their influence on the physiolog- 
 ical condition of the plant or directly, through their action on the 
 cells of the separation zone. Most investigators, except Wiesner, ex- 
 
1918 1 Kendall: Abscission of Flowers and Fruits in Solanaceae 361 
 
 press the opinion that atmospheric factors work directly in causing 
 "spontaneous" abscission, although offering, so far as I can see, no 
 evidence for this view. Fitting states that the external influence 
 acts directly in most cases, but that the indirect action is apparent in 
 forms which must build a separation layer before fall can occur. 
 In regard to the action of injury, it seems to be the opinion of most 
 investigators (Hannig, Bacquerel, Gortner and Harris)-- -that the 
 stimulus acts indirectly by interfering in some way with such 
 important physiological processes as transpiration, respiration, or 
 assimilation. On the other hand, if abscission is sometimes a semi- 
 tropistic phenomenon, as Fitting has suggested, it is evident that 
 injury may act directly in causing flower-fall. 
 
 TECHNIQUE 
 
 The results noted below were obtained largely from the examina- 
 tion of microscopic preparations made by the paraffin method, 
 although this method was supplemented by free-hand sections mounted 
 in water. In investigating the condition of the pedicel in some species 
 (Datura sp., Petunia sp. and several species of Nicotiana) only free- 
 hand sections were examined. For most microchemical studies fairly 
 thick, free-hand sections are preferable. The material for sectioning 
 in paraffin was killed and fixed in various concentrations of the 
 chromo-acetic series and dehydration and infiltration were, in general, 
 carried on very slowly. The free-hand sections were mounted in water 
 without killing. 
 
 In cutting longitudinal sections of any kind all the pedicels were 
 oriented so that the sections were cut parallel to the main stem of the 
 inflorescence, in the plane formed by the pedicel and stem taken 
 together. In studying the- histology of the pedicel and the cytology 
 of the separation layer and in studjdng the method of cell separation, 
 these longitudinal sections were supplemented by cross sections in 
 series through the base of the pedicel. It was impossible to cut very 
 thin, longitudinal sections in paraffin without crushing or breaking 
 the cells ; most of these sections therefore were cut from 10/x to 15/* in 
 thickness. For a similar reason, it was found necessary to cut thick 
 sections (20/x to 25/*) of the pedicels of fruits in which mechanical tissue 
 had developed. It was possible, however, to cut excellent paraffin 
 sections from 5/x, to 1^ in thickness in cross-section or longitudinally 
 through the small cells of the separation zone. Since the cells of the 
 
362 University of California Publications in Botany [VOL. 5 
 
 separation zone are very small, not much could be determined in 
 regard to the dissolution of cell walls by means of thick, free-hand 
 sections. The best results along this line were obtained from the thin 
 paraffin sections of the separation zone, although in order to show the 
 cell wall in its normal thickness it was necessary to use the free-hand 
 sections. As a supplement to these sections, several points of interest 
 were brought out by washing off the isolated cells from the end of 
 freshly abscissed pedicels and mounting them for microscopic exam- 
 in.ation. 
 
 In most of the work the paraffin sections were stained in safraniii 
 and Delafield's haematoxylin. The free-hand sections were generally 
 mounted in water and stained in iodine. In special instances other 
 stains were used. Thus, in testing for chemical differences in the cell 
 walls of the separation cells, several other stains, such as erythrosin, 
 eosin, Bismark brown, gentian violet and Ruthenium red were used. 
 It was found that for demonstrating the dissolution of cell walls 
 aqueous methylene blue was an excellent stain to use. This stain was 
 allowed to act overnight and the sections destained slightly in alcohol. 
 Methylene blue was also an excellent stain for the isolated cells ob- 
 tained as noted above. By fixing these cells to the slide with albumen 
 fixative and staining with this stain, the thin membranous wall sur- 
 rounding the protoplast can be distinctly seen. 
 
 Various methods, such as subjecting inflorescences to illuminating 
 gas and mechanical injury, were used to bring about abscission. The 
 best results were obtained in cases where abscission was induced by 
 inserting shoots under a bell-jar containing from 1.5 per cent to 
 3 per cent illuminating gas. By using illuminating gas in this way 
 and by taking sections of the pedicels at intervals it was possible to 
 determine just when the first signs of abscission appeared in a certain 
 percentage of gas. This time was definitely determined for certain 
 species so that it was possible to get material killed and fixed at any 
 desired stage in the process of abscission. It was found that the best 
 results were obtained by killing and fixing the pedicels at about the 
 time when abscission was known to be commencing. 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 363 
 
 HISTOLOGY AND CYTOLOGY OF THE PEDICEL 
 
 1. HlSTOLOGICAL AND CYTOLOGICAL CONDITIONS OF THE 
 
 MATURE PEDICEL 
 a. NICOTIANA 
 
 The vascular system in Nicotiana, as in all the other genera 
 examined, is characterized by intraxylary phloem. Nicotiana differs 
 slightly from all others in that the xylem seems in cross-section 
 to be composed of a continuous ring of radial strands of tracheae 
 rather than composed of a broken ring of distinct bundles. When a 
 branch of the vascular system (fig. 1, a) containing twenty to thirty 
 xylem strands is given off to the pedicel, it assumes the shape of a 
 crescent in cross-section, with the opening of the crescent on the ven- 
 tral side. A short distance distal to the groove which marks the sep- 
 aration zone (fig. 1,6), the crescent closes and throughout the 
 remainder of the pedicel the vascular system forms a complete cylin- 
 
 c - 
 
 Fig. 1. Diagram of pedicel of Nicotiana 
 
 a vascular system. 
 b separation zone. 
 c pedicel cortex. 
 sc stem cortex. 
 e epidermis. 
 
 f chlorophyllous tissue. 
 g groove. 
 h separation layer. 
 p pedicel pith. 
 
364 University of California Publications in Botany [VOL. 5 
 
 der. The pith and cortex (fig. 1, p and c) are composed of large 
 parenchyma cells which in the cortex are two or three times as long 
 as wide, but in the pith are more nearly isodiametric. There is no 
 mechanical tissue to be found in the floral pedicel but, as will be noted 
 in more detail later, wood fibres are formed as soon as the fruit begins 
 to develop. The epidermis of the pedicel (fig. 1, e) is typical but with 
 a poorly developed cuticle, especially in the groove (fig. 1, #), where 
 the cells are also much reduced longitudinally. Beneath the epidermis 
 is a layer of small cells with very large intercellular spaces and an 
 abundance of chloroplasts (fig. I,/). This tissue stops a short dis- 
 tance proximal to the separation zone and does not continue in the 
 pedicel. The layer of collenchyma which is commonly found in cer- 
 tain species just beneath this chlorophyl tissue is entirely absent in 
 Nicotiana, or at least is very poorly developed. 
 
 Corresponding with the general region of the groove is an area of 
 medullary and cortical cells which are smaller than corresponding 
 cells on either the proximal or distal side of the groove. This region 
 of small cells is homologous with the separation zone (fig. 1, &) and it 
 extends across the base of the pedicel. The smallest cells are in the 
 center of the region, in a plane with the bottom of the groove, and 
 grade in size to the larger cells of the pith and cortex on either side 
 (plate 49, fig. 1). The zone of small cells is ten to fifteen tiers of 
 cells thick on the dorsal side but is wider on the ventral side, where it 
 spreads out into the large area of storage cells found in the axil of 
 the pedicel. The separation layer (fig. 1, h) is located five to seven tiers 
 of cells distal from the bottom of the groove. Hanning reports this 
 layer as occurring at the tip of the pedicel in Nicotiana Langsdorffii, 
 but in all my experiments on two varieties of this species I find separa- 
 tion invariably occurring at the base of the pedicel in the position 
 described above. All the species and varieties of Nicotiana examined 
 show a structure of the pedicel corresponding with the above descrip- 
 tion except that in some varieties, as in those of N. Bigelovii, the sep- 
 aration zone is much thinner on the dorsal side. In such cases it is 
 also noted that the groove is poorly developed. 
 
 The cells of the separation layer are in no way morphologically 
 differentiated from those making up the remainder of the separation 
 zone. Indeed, any cell of the zone seems capable of functioning as a 
 separation cell. The separation cells are smaller than normal cortical 
 cells and spherical in shape except in the vascular bundles, where they 
 do not seem to.be differentiated in size and are elongated parallel to 
 the longitudinal axis of the pedicel. The cell walls are slightly thicker 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanac&ae 365 
 
 than the walls of normal cortical cells, especially at the corners, thus 
 giving the tissue a somewhat collenchymatous appearance. The small- 
 est cells more proximal show this collenchymatous nature more strik- 
 ingly than do the others. No difference in chemical composition could 
 be detected, by means of microchemical tests using caustic potash, 
 sulfuric acid, nitric acid, and various stains, between the_cell walls of 
 the separation cells and walls of other cortical cells. Other tests, how- 
 ever, indicated a difference in the nature of the cell contents in the 
 two types of cells. Iodine frequently indicates the presence of starch 
 in these cells and also colors the protoplasts a darker brown than in 
 normal cells, showing that the separation cells are rich in protoplasm. 
 The amount of starch in the cells, however, was found to be extremely 
 variable, ranging from a total absence of starcli to an abundance of 
 it. Iodine green imparts to the protoplast of the separation cells a 
 deep blue color in contrast with other cortical cells, which are not 
 colored by this stain. The blue reaction is most prominent where the 
 separation layer crosses the phloem. Other cells which react in the 
 same way to this stain are the sieve tubes and companion cells and 
 the storage cells in the axil of the pedicel. 
 
 b. LYCOPERSICUM 
 
 Conditions in Lycopersicum differ in certain respects from those 
 existing in Nicotians. In the former the separation zone (fig. 2, a) 
 seems to be located at the middle of the pedicel 
 and is marked externally by a swelling, as well 
 as by the groove of the type already noted as 
 characteristic of the pedicel of Nicotiana. This 
 groove in the tomato is very deep (plate 53, 
 fig. 1), reaching fully half the depth of the 
 cortex, and is, furthermore, of about the same 
 depth all the way round, differing in this 
 respect from Nicotiana, where the groove is 
 absent or poorly developed on the ventral 
 side. The vascular system in Lycopersicum 
 (fig. 2, &), in contrast with the condition in 
 Nicotiana, is composed of scattered bundles 
 of xylem which in this case do not form a 
 crescent proximal to the groove but are in the 
 form of a complete cylinder throughout the 
 entire pedicel. Beneath the epidermis (fig. 
 
 r ~C 
 
 Fig. 2. Diagram of pedicel 
 
 of Lycopersicum 
 a separation zone, 
 ft vascular system. 
 c epidermis. 
 <i separation layer. 
 e pith. 
 f chlorophyl-bearing 
 
 tissue. 
 g collenchyma 
 
366 University of California Publications in Botany [ V OL. 5 
 
 2, c) is the chlorophyl-bearing region of the cortex (fig. 2, f), such 
 as occurs in Nicotiana, but in this case the tissue continues in the 
 pedicel distal to the groove. Beneath this chlorophyl-bearing tissue 
 is a layer of well-developed collenchyma (fig. 2, g) which however 
 does not continue in the pedicel distal to the groove. The separation 
 layer (fig. 2, d) consists of three to six tiers of cells and is located 
 in a plane with the groove, differing in this respect from Nicotiana, 
 where it is located a short distance distal to the groove. Correspond- 
 ing to the condition in Nicotiana, the chief characteristic of the separa - 
 tion cells is their small size, spherical outline and active physiological 
 condition. 
 
 o. OTHEE GENERA OF THE SOLANACEAE 
 
 The condition of the pedicel, so far as the histology of the separa- 
 tion zone is concerned, was examined in several other species, a list 
 of which is given below : 
 
 Solanum jasminioides Oestrum fasciculatum 
 
 Solanum tuberosum lochroma tuberosa 
 
 Solanum verbascifolium Datura sanguineum 
 
 Solanum umbelliferum Salpichrora rhomboidea 
 
 Solanum nigrum Petunia hybrida 
 
 Solanum marginatum Salpiglossis sinuata 
 
 Lycium australis 
 
 The general condition of the pedicel of Datura sanguineum and 
 Petunia hybrida is worth describing in some detail. The tissues of 
 plants of D. sanguineum are more or less herbaceous in nature, large- 
 celled and somewhat succulent throughout. The chlorophyl-bearing 
 tissue which, in striking contrast with the condition in Nicotiana and 
 Lycopersicum (figs. 1 and 2), is continuous over the separation zone, 
 is composed of two rows of small, spherical cells just beneath the 
 epidermis. Except for a layer of collenchyma, whose much elongated 
 cells extend the entire length of the pedicel and thus continue the col- 
 lenchyma through the separation layer, the cortex and pith are com- 
 posed of more or less isodiametric, thin-walled cells. Floral abscission 
 is as common in this species as it is in Nicotiana. The flowers are 
 very large and furnish excellent material for a study of the cytology 
 of abscission. Unfortunately not a sufficient number of flowers could 
 be obtained to make possible any detailed study of this genus. It was 
 noticed, however, that there is no region of small cells at the base of 
 the pedicel within which separation occurs and that the separation 
 cells are identical in size and shape with those on either side among 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 367 
 
 which separation does not occur. The separation layer here is located 
 about 8 mm. distal to the base of the pedicel, with absolutely no ex- 
 ternal indication of its position. Microchemical tests, which in Nioo- 
 tiana gave different reactions in the case of the separation zone and in 
 the case of normal cortical cells, here fail to show any corresponding 
 condition of differentiation. 
 
 Abscission has never been found to occur in Petunia or Salpiglossis, 
 so that it is of interest to examine the histological condition of the 
 base of the pedicel in these two species. They are practically identical 
 with regard to the structure of the pedicel, so that the description 
 given below can be taken as applying to both genera. The cortical 
 cells of the pedicel pass into those of the stem without any groove or 
 small-celled region. On the ventral side, however, is the region of 
 small cells in the axis of the pedicel, which is more or less common to 
 all flowers. The tissues of Petunia are not so soft and succulent as 
 those of Datura, Nicotiana, and Lycopersicum. They tend rather to 
 be dry and tough. The cells in the cortex and pith are also not so 
 nearly isodiametric as in Datura, but are much elongated in a direc- 
 tion parallel with the long axis of the pedicel. 
 
 The condition in the other species mentioned above will be given 
 only a general description. Abscission occurs in all the other species 
 except Salpichrora and Lycium which, however, do not differ, in 
 respect to the histology of the base of the pedicel, from any of the 
 others. Solanum tuberosum resembles Lycopersicum. All the other 
 species are similar in regard to the structure of the separation zone. 
 There is in every case a general region of small cells extending 
 across the base of the pedicel where the separation layer occurs. 
 
 3. DEVELOPMENT OF THE SEPARATION ZONE IN Lycopersicum 
 AND Nicotiana 
 
 a. LYCOPEESICUM 
 
 The development of the separation zone could be followed better 
 in Lycopersicum than in Nicotiana because in the former the zone is 
 not so close to the main axis of inflorescence. The problem here 
 resolves itself into an effort to determine, by means of longitudinal 
 sections of very young pedicels, how early in the development of the 
 flower the groove and the differentiation in cell size of the separation 
 cells appear. It was found that the development of the separation zone 
 indicates the method by which the groove and differentiation in cell 
 
368 
 
 University of California Publications in Botany [ V OL. 5 
 
 size originate. The groove is fairly well developed (fig. 5) in young 
 buds whose corolla is only 3 mm. in length, but is not so deep as in 
 older buds. The cells of the separation zone at this stage are smaller 
 than cells on either side, but the difference is not so prominent as in 
 older flowers. In very small buds whose corolla is only 1 mm. in 
 length or whose calyx is only 2 mm. long, the groove is just beginning 
 to appear (fig. 4). In buds below this size (fig. 3) no groove or 
 differentiation in cell size can be detected. Abscission can occur in 
 these early stages, before the groove or differentiation in the size of the 
 separation cells has appeared, as well as at any other stage. In these 
 early stages the radial diameter of the cortex is much less, as com- 
 pared with that of the pith, than in older flowers. It is evident, 
 therefore, that the cells of the separation zone are small because they 
 retain their original small size while the rest of the cortical cells 
 increase in size. The fact that the groove is formed makes it probable 
 that there have been few cell division, or none, in the separation zone 
 of the cortex during the development of the bud. It was observed, 
 however, that the cells of the separation zone in the pith retain their 
 meristematic nature for a considerable period during the development 
 
 r 
 
 / 
 
 Fig. 3 
 
 Fig. 4 
 
 Fig. 5 
 a separation zone 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 369 
 
 of the flower. They are at this time rectangular in shape, elongated 
 perpendicularly to the long diameter of the pedicel, and arranged in 
 longitudinal rows. In later stages, however, w r hen the flower is at 
 anthesis, or the fruit is forming, these cells have rounded up and 
 become irregularly arranged, thus leaving rather large intercellular 
 spaces. 
 
 fc. NICOTIANA 
 
 The separation zone develops in Nicotiana much as it does in 
 Ly coper sicum. It was observed in very young buds calyx 2 or 3 mm. 
 or shorter that no groove was present. In buds larger than these, 
 the groove and small size of the separation cell is apparent, appearing 
 first on the dorsal side of the pedicel. It is evident that in both these 
 genera the groove and area of small cells are explained in the same 
 way, i.e., by the fact that the normal cortical cells increase in size 
 faster than do the cells of the separation zone. Since in both genera 
 abscission can occur even before differentiation of any kind appears at 
 the base of the pedicel, it is evident that the groove and small-celled 
 region do not necessarily bear any relation to abscission. This state- 
 ment is borne out by the fact that in Datura neither the groove nor 
 the area of small cells is present and in Nicotiana separation occurs a 
 short distance distal to the groove. 
 
 c. CONCLUSIONS FBOM THE STUDY OF THE DEVELOPMENT OF 
 THE SEPAEATION ZONE 
 
 In view of the above discussion it is clear that the separation layer 
 in Ly coper sicum, Nicotiana, Datura, and probably in the other genera 
 noted, originates according to the first method, a, proposed by Kubart 
 (cf. page 350). That is to say, the separation layer represents merely 
 a portion of the primary meristem which retains its original physi- 
 ological capacities. 
 
 4. INCREASE IN SIZE AND DEVELOPMENT OF MECHANICAL TISSUES IN 
 THE PEDICEL OF Nicotiana AND Lycopersicum 
 
 There is a marked increase in the size of the pedicel in both 
 Nicotiana and Lycopersicum during the development of the fruit. It 
 was found that during this development the diameter of the pith 
 remains about the same, the actual increase in size being almost 
 entirely confined to the cortex (cf. figs, 3, 4, and 5). This increase in 
 the diameter of the cortex in the pedicel of Nicotiana is due, in the 
 first place, to an increase in the size of the original cortical cells, 
 
370 University of California Publications in Botany [VOL. 5 
 
 which in average cases measured about 20/x in diameter in the flower 
 and about 40/x, in the fruit. In the second place, it is due to four or 
 five divisions of the cambium layer. This second factor in the increase 
 in size of the pedicel becomes evident when a count is made of the cells 
 between the phloem and tracheae, the result giving approximately six 
 cells in the flower and eleven in the fruit. 
 
 The increase in size of the pedicel of Lycopersicum, which is much 
 more prominent than the increase in Nicotiana, can be explained in 
 the same manner. In the former the increase in size, which in this 
 case takes place almost entirely distal to the groove, may proceed to 
 such an extent that the diameter of the pedicel of the fruit is two or 
 three times that of the flower at anthesis. A measurement of the 
 cortical cells in cross-section gave on the average 10//, in the flower and 
 28/x in the fruit. In this case only two or three divisions of the 
 cambium occur; the cells resulting immediately show lignification. 
 
 The next subject of consideration is the development of mechanical 
 tissue in the pedicel of Nicotiana and its relation to abscission. It will 
 be remembered that there was no mechanical tissue noted in the 
 pedicels of buds and flowers. Parallel with the development of the 
 fruit, however, a continuous ring of mechanical tissue appears in the 
 xylem of the pedicel. This mechanical tissue is evidently the result 
 of a gradual lignification of the cells of the cambium and the outside 
 portion of the xylem parenchyma. There is thus formed a continuous 
 sheath of what may best be called wood-fibre tissue, in the form of a 
 cylinder just outside the tracheal elements. These mechanical elements 
 first appear in the tissues of the pedicel five or six days after anthesis, 
 but since the lignification in these more distal tissues is merely the 
 result of the spreading upwards of the lignification in the older parts 
 of the plant, this period depends somewhat on the position of the 
 flower on the inflorescence. It was noticed in Nicotiana that the 
 wood-fibre tissue develops on both sides of the separation zone before 
 appearing in the latter, but in time it becomes continuous through 
 the separation layer. By a lignification of the cells between the two 
 points of the crescent of wood in the separation zone, there is also 
 a slight tendency to close this crescent on the ventral side. 
 
 Since abscission has not been observed to occur in lignified cells, 
 the question at once arises whether the tough sheath of lignified cells 
 which continues through the separation layer could hold the fruit on 
 the stem even after actual abscission had occurred. Upon looking 
 over any large number of plants in the field it will at once be evident 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 371 
 
 that such a condition of affairs very often exists. It will be found in 
 many cases, especially on older plants, that although abscission has 
 occurred in the cortex, as evidenced by the presence af a white, 
 powdery substance at the base of the pedicel, the capsule is yet firm 
 on the stem. Indeed, in certain hybrid tobaccos it is common to find 
 most of the capsules in this abscissed condition. The fruit is supported 
 in these cases by the tough mechanical elements of the wood, which 
 also prevent the breaking of the tracheae and protect the intraxylary 
 phloem. In the pith the tissues may be in a somewhat abscissed con- 
 dition, but since there is no way for these cells to escape through the 
 sheath of wood they remain for some time in position before finally 
 collapsing. 
 
 The development of mechanical tissues takes place in Lycopersicum 
 in much the same manner as in Nicotiana but with the distinct differ- 
 ence that in the former the wood-fibre tissue does not become con- 
 tinuous through the separation layer. That is to say, in the tomato 
 a break is .left in the mechanical tissue in a plane with the bottom of 
 the groove. It is evident here that abscission would cause fall of the 
 fruit in any stage of its development, although in this case it happens 
 that abscission very rarely occurs after two or three days past anthesis. 
 A condition resembling this one in the tomato was observed in other 
 berry-forming species of the Solanaceae such as Oestrum fasciculatum 
 and Solanum verbascifolium, which often drop their immature fruits 
 by abscission. Abscission, however, very seldom occurs in mature 
 berries of these species, the fruit generally falling away from the 
 receptacle above the calyx. 
 
 THE PROCESS OF ABSCISSION 
 
 1. GENERAL DESCRIPTION OP THE PROCESS IN SEVERAL GENERA 
 
 a. NICOTIANA 
 
 The process of abscission in all the species of Nicotiana investi- 
 gated conforms to the usual type involving separation and isolation of 
 cells. Further details of the process were briefly discussed in a pre- 
 liminary paper (Goodspeed and Kendall, 1916) for certain F : species 
 hybrids of Nicotiana. It was there noted that cell separation starts in 
 the dorsal side of the pedicel, in the cortex a short distance distal to 
 the groove (pi. 49, fig. 1) and spreads from this point around to the 
 ventral side. The first external indication seems to be a bulging of 
 
372 University of California Publications in Botany [VOL. 5 
 
 the epidermis (pi. 49, fig. 2) over the tissue in which the process is 
 taking place. Simultaneously with the start of abscission in the 
 cortex, the process apparently originates independently in the pith 
 (pi. 50, fig. 1). It was further noted that the number of cells con- 
 cerned in the process, as a general rule, is greater in the hybrids than 
 in their parents and also that this is true of ' ' automatic ' ' as compared 
 with ''spontaneous" abscission. Just beneath the epidermis the cells 
 involved in separation were reported as being from five to ten tiers 
 thick, but as the process approached the vascular tissue the separation 
 layer was evidently reduced in thickness to not over one or two tiers 
 of cells (pi. 52, fig. 1). In the pith a more or less spherical mass of 
 cells is involved (pi. 50, fig. 1). When the separation is completed 
 the flower may remain in position for some time, until the epidermis 
 and tracheal elements are broken by some mechanical agency. 
 
 The exposed separation surface of the pedicel was stated to be 
 convex in outline and slightly notched at the tip. Upon closer exam- 
 ination the surface itself was seen to be composed of the protruding, 
 rounded ends of cells with here and there completely isolated cells and 
 broken ends of spiral tracheae. These isolated cells are apparently 
 normal and do not markedly differ in form, size, or in the nature of 
 their cell inclusions from the same cells before separation. The 
 exposed surface of the attached portion of the pedicel is similar in 
 appearance to that of the detached portion, but is more or less flat 
 in outline. After separation the cells on this surface collapse and 
 probably act as a protective layer. 
 
 Following the observations recorded above, which had to do largely 
 with flower-fall in the F^ species hybrids, a number of species have 
 been investigated in an effort to determine whether or not their mode 
 of abscission differs from that already described. 
 
 It may be noted at the start that no marked exceptions were found 
 to the previously described condition, although at least two stages in 
 the process of abscission have been found to be subject to considerable 
 variation. The first of these stages has to do with the place of origin 
 of the abscission process itself. An independent origin in the pith 
 has been demonstrated to occur in a large number of species and 
 occasionally it was found that the first evidences of abscission could 
 be detected here before any similar evidences appeared in the cortex. 
 Again, it was found in most species that cell separation starts first in 
 the ventral cortex although other places of origin were found in 
 several cases. Thus, in Nicotiana Tabacum "Maryland" and F, TIM. 
 
Kendall: Abscission of Flowers and Fruits in Solanaceae 373 
 
 for example, the process originates on the ventral side and may even 
 spread through the large area of storage cells in the axil of the pedicel 
 before reaching the dorsal side. The distance distal from the bottom 
 of the groove at which separation appears is also subject to variation. 
 This variation, however, is not typical of certain species, since it may 
 occur at different times in the same species, evidently as a result of 
 an abnormal stimulation to abscission. 
 
 The second part of the process subject to variation has to do with 
 the amount of tissue that may be concerned in actual cell separation. 
 Abscission first becomes complete in a narrow plane between two or 
 three tiers of cells across the pedicel and the flower can be easily 
 shaken off at that time. If, however, the flower remains on the stem, 
 and is kept turgid by the water rising in the unbroken trachea?, cell 
 separation spreads more and more widely through the tissues of the 
 pedicel, especially in the pith and cortex. It is the extent to which this 
 spreading normally proceeds that varies in the different species. When 
 the process has spread to a considerable extent, a white ring formed 
 by the isolated masses of cells can be seen with the naked eye at the 
 base of the pedicel and a casual inspection indicates that the amount 
 of this white substance varies in the different species. In most 
 hybrids, except F x H179, there is more spreading in normal abscission 
 than in pure species. In Nicotiana quadrivalvis, N. Bigelovii, and 
 other similar species in which abscission very seldom occurs, no spread- 
 ing takes place. Spreading, however, occurs to a remarkable extent in 
 N. Tabacum "Maryland." 
 
 ft. LYCOPEBSICUM 
 
 We may say that, in general, abscission in Ly coper sicum corre- 
 sponds to that in Nicotiana and that the main points of distinction 
 between the two arise only from the original differences in the separ- 
 ation zones (cf. page 364). In addition, attention must be called to 
 the fact that quite frequently, in individual plants of the tomato, no 
 true abscission occurs in normal flower-fall. In these cases the flower 
 seems to be detached from the plant by a process which compares 
 closely with that called exfoliation. There is no active cell separation 
 and the flower simply wilts and dries back to the groove, where it 
 hangs until broken off by some mechanical agency. The first indica- 
 tion of the process is the loss of chlorophyl in the pedicel, which 
 gradually turns yellow, commencing at the tip and spreading proximal 
 to the separation zone. It is possible that most of the flower-fall 
 
374 University of California Publications in Botany [ V L - 5 
 
 noticed by agriculturists is of this type. Quite often, however, true 
 abscission and this second type of flower-fall may both be found 
 operative in the same plant or even in the same flower. "Spon- 
 taneous" flower-fall in the tomato is, of course, of the true abscission 
 type. 
 
 Corresponding with the condition in Nicotiana, true abscission in 
 Lycopersicum is seen to originate frequently in the pith. At any rate, 
 the process goes on here independently of that in the cortex, since the 
 final break is through the tracheae and epidermis. Furthermore, separ- 
 ation takes place in a plane with the bottom of the groove (pi. 53, fig. 
 2) whereas, in Nicotiana, it takes place a short distance distal to the 
 groove. Separation may at first take place between only two tiers of 
 cells (pi. 53, fig. 2), but in time the process may spread until three 
 or four tiers become involved in separation. However, there is no 
 spreading of the process to a large number of cells, as is frequently 
 seen in Nicotiana, so that one very seldom finds the white powdery 
 substance at the point of separation. Also in contrast with the con- 
 dition in Nicotiana, there is, as abscission progresses, no bulging of 
 the epidermis which instead soon breaks in the bottom of the groove. 
 Separation in the tomato takes place in such a way as to give the 
 exposed separation surfaces the same general shape after abscission as 
 in Nicotiana, that of the detached portion of the pedicel being convex 
 and that of the remaining portion slightly concave. 
 
 c. DATURA 
 
 Conditions in Datura differ strikingly from those in the two 
 species described above. This would be expected when one considers 
 the great differences in the structure of the separation zones (cf. 
 page 365). In Datura there is the usual chlorophyl-bearing tissue, 
 which consists of two rows of small, perfectly isodiametric cells with 
 large intercellular spaces, just beneath the epidermis. It will be 
 remembered from the description on page 365 that this tissue in Datura 
 continues the entire length of the pedicel and therefore, in contrast 
 with the condition in Nicotiana and Lycopersicum, extends through 
 the separation zone. The first sign of abscission is the maceration of 
 this tissue as indicated by the appearance of a white color under the 
 epidermis. The latter may as a result become detached from the tissues 
 of the cortex for a distance of 2 cm. or more along the base of the 
 pedicel. This is soon followed by a break over the separation layer 
 and a curling back of the epidermis on either side, with most of the 
 chlorophyl-bearing cortical tissues still attached to its inner surface. 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 375 
 
 After the break in the epidermis separation continues in the layer 
 of collenchyma just beneath. The cells of the collenchyma layer, 
 which are much elongated parallel to the long axis of the pedicel (five 
 to eight times as long as wide), separate for a distance of about 0.3 
 mm. up and down the pedicel, involving only a few tiers of cells. It 
 is evident that the cells of this tissue separate without difficulty, 
 although not by any means as freely as the small spherical cells de- 
 scribed above. The large, isodiametric, parenchyma cells of the cortex 
 separate for a distance of 2 or 3 mm., involving many tiers of cells. 
 The cells of the starch sheath, which are small and spherical, separate 
 for a distance of 1 cm. or more, thus causing a longitudinal cavity to 
 be formed just outside of the vascular bundles. In the latter, separa- 
 tion involves only two or three tiers of cells. Separation originates 
 and continues in the pith independent of the process in the cortex, 
 but involves about the same number of cells as in the parenchyma of 
 the latter tissue. "When separation has thus become complete, the 
 weight of the flower is very often sufficient to break the tracheag and 
 cause the flower to fall to the ground. 
 
 Several important facts are brought out by this examination of 
 abscission in Datura. In the first place, it shows that floral abscission 
 can take place without any structure which might possibly be inter- 
 preted as a morphologically differentiated separation layer. In the 
 second place, it indicates that cell separation is possible in several dif- 
 ferent types of living cells. It also shows that separation takes place 
 more readily in small cells than in large ones and more readily in 
 isodiametric cells than in elongated ones. The theory that the separa- 
 tion layer is not a morphologically differentiated structure, but repre- 
 sents a physiological condition (Lloyd and Loewi), could certainly be 
 well applied in this case. 
 
 d. OTHER GENEEA 
 
 The process of abscission in the other species listed on page 365 is 
 essentially the same throughout. No indications were noted of cell 
 divisions or elongations accompanying abscission. Separation is 
 brought about by means of a separation of small and active cells 
 located in the general region at the base of the pedicel. In all these 
 forms the separation surface of the pedicel is convex in outline, so 
 that the separation layer must lie in more or less of a crescent in the 
 stem at the base of the pedicel. The main difference between these 
 forms and the three that have been described in detail above is found 
 
376 University of California Publications in Botany L V L - 5 
 
 in the fact that in the former, with the exception of Solatium tuber- 
 osum, separation occurs in the stem, at the very base of the pedicel, 
 whereas in the latter three it occurs through the pedicel a varying 
 distance from the base. 
 
 2. METHOD OF CELL SEPARATION 
 
 a. GENERAL REMARKS 
 
 It will be remembered that two theories have been proposed to 
 account for the cell separation that is responsible for abscission. 
 First, it is conceivable that cell separation may be caused by an 
 increase in cell turgor, which causes the cells to round up and pull 
 apart without any change taking place in the chemical nature of the 
 middle lamella. Second, cell separation may be caused by a chemical 
 dissolution of the middle lamella with or without an increase in cell 
 turgor. The main difference between the two theories is that the 
 second, in contrast with the first, maintains that chemical alteration 
 of the middle lamella is always necessary before abscission can occur. 
 The first theory gains support from the work of Pitting and the second 
 from the work of Hannig, Lee, Strasburger, and Lloyd. Wiesner, 
 Kubart, and Loewi believe that cell separation takes place by the 
 action of both factors but that either factor may at times be the more 
 important. 
 
 I. CYTOLOGICAL CHANGES ACCOMPANYING ABSCISSION 
 
 It was stated in a preliminary discussion (Goodspeed and Ken- 
 dall, 1916) first, that no indication of cell divisions or elongations 
 were observed accompanying abscission, and, second, that no evidence 
 of the dissolution of the middle lamella had at that time been obtained. 
 The first statement has been corroborated in that, during all the later 
 experiments, no divisions or elongations have been observed in any of 
 the described species. The dissolution of the primary cell membrane, 
 however, because of more exact knowledge of the proper time to take 
 sections and of more successful staining methods, has been fairly well 
 established. 
 
 The main problem here was to determine by the use of various 
 stains whether or not the primary and secondary cell membranes of 
 the separation cells stain differently in the early stages of abscission 
 than under normal conditions. This was a point which was found 
 very difficult to determine, principally because of the fact that the 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 377 
 
 separation cells are, comparatively speaking, very small, but also be 
 cause of the fact that the walls of these cells fail to show any strati- 
 fication. 
 
 Iodine, Delafield's haematoxylin, Ruthenium red, Bismark brown, 
 methylene blue, erythrosin, and eosin were used with little success in 
 most cases. By using iodine, however, just as abscission is known to be 
 commencing, a white streak may be seen across the section in the region 
 of the separation layer. Upon careful examination it was decided 
 that this white streak was due to the failure of most of the cell walls 
 in the separation layer to take the stain. Although it is probable 
 that with more careful examination the other stains mentioned above 
 would give similar results, it was found that methylene blue was the 
 only stain with which anything definite could be established. If a 
 thin longitudinal section cut in paraffin as abscission is known to be 
 starting, and stained in methylene blue, is examined (cf. page 361), it 
 will be found that the walls of those cells in which separation is about 
 to occur have remained almost entirely unstained. The protoplasts in 
 these cases seem to be surrounded only by the thin tertiary mem- 
 branes, between which is a streak of colorless material of varying 
 width (pi. 51). Cell walls where separation is not expected to occur, 
 however, stain a dark blue throughout in the normal manner. 
 
 An examination of freshly isolated cells washed oft 2 from the end 
 of an abscissed pedicel shows that these cells are still turgid and 
 active. It was impossible to determine whether these cells had in- 
 creased in size, as compared with the size of similar cells before abscis- 
 sion, but it is evident that the increase, if any, had not been very 
 great. The cells still contain their large nuclei, and occasional starch 
 grains, and show after isolation no signs of degeneration even after 
 several hours in water. In addition, these isolated cells appear to have 
 retained their original shape. In the collenchyma of Datura the cells 
 are from five to eight times as long as wide, and yet these cells retain 
 their original shape when isolated, as a result of the dissolution of 
 the middle lamellae. This isolation has evidently not been complete, 
 since large masses of cells are seen still attached to each other. It is 
 noticed that in all cases the protoplast is surrounded by an extremely 
 thin membranous wall (pi. 52, fig. 3). It is also frequently noticed 
 that the protoplast seems drawn away from the cell wall as if plas- 
 molysis had occurred. It is possible that this appearance may be due 
 simply to the gathering together of granules and the denser portion 
 of the protoplasm in the center of the cell. 
 
378 University of California Publications in Botany [VOL. 5 
 
 c. EXPEEIMENTAL EVIDENCE FOE THE DISSOLUTION OF 
 THE MIDDLE LAMELLA 
 
 It is supposed that the middle lamella, or primary cell membrane, 
 is largely composed of calcium pectate, a calcium salt of pectic acid 
 which has been given the general name pectose. The secondary cell 
 membranes probably contain a larger proportion of cellulose with 
 the pectose than is present in the primary membranes. This pectose, 
 which is of course insoluble in water, is disorganized by a process of 
 hydrolysis to form pectin. The pectin, which is a colorless mucilagi- 
 nous substance, is readily soluble in water but is precipitated along 
 with the proteids and enzymes of the protoplast by the addition of 
 alcohol. Thus, if a water extract is made from separation zones dur- 
 ing the first stages of abscission, one would expect to get a solution of 
 several substances, among which would be the pectin produced by the 
 dissolution of the pectose in the primary cell membranes. It might 
 be expected that the amount of precipitate obtained from this extract 
 with alcohol would be greater, provided the amount of other sub- 
 stances remained the same, than the amount of precipitate obtained 
 in a similar manner from separation zones in which there had been 
 no abscission and in which no pectin had been formed. Whether or 
 not the increase in the amount of precipitate is due to the added 
 pectin cannot of course be proven without actual chemical analysis, 
 and such an analysis would be difficult because of the very small sam- 
 ples of material obtainable. However this may be, any difference in 
 the amount of precipitate would be of interest. 
 
 This experiment and the two which follow are, as far as I have 
 been able to determine, the first of their kind. Apart from this fact, 
 their chief value probably lies in the fact that they suggest a line of 
 investigation which, if carried on in more detail and with better 
 facilities, will undoubtedly lead to important conclusions. These 
 experiments were, however, carried on with as much care as possible 
 and since the results of duplicate tests are in agreement, they give, as 
 far as they go, dependable results. 
 
 After several experiments indicating the results given below, the 
 following test experiment was performed: 
 
 Experiment 1. Two water extracts of equal concentration were 
 made from the lots of material. Lot A contained 200 small pieces of 
 the pedicel in which the separation zone was located and in which 
 abscission had started. Lot B contained an equal weight of a similar 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 379 
 
 number of pedicels in which no abscission had started. The extracts 
 were made up to 10 cc. and the precipitate obtained with 60 cc. of 
 95 per cent alcohol. The precipitate weighed in the two lots : 
 
 A 996 nig. 
 
 B : 903 mg. 
 
 One of the preliminary experiments performed with~~a "weaker 
 alcohol gave results which may or may not be of considerable impor- 
 tance. In this experiment a light, almost invisible precipitate formed 
 in A and no precipitate in B. Whether or not the pectins precipitate 
 in lower percentages of alcohol more readily than the other substances 
 I have been unable to determine. At any rate, the precipitate in this 
 case felt slimy and mucilaginous to the touch and might well have 
 been the precipitated pectin approximately pure. 
 
 d. EVIDENCE FOE INCEEASE IN TUEGOE 
 
 It was stated along with other conclusions in the preliminary paper 
 (Goodspeed and Kendall, 1916) that from the evidence at that time 
 available it was probable that cell separation is caused merely by an 
 increase in cell turgor, and throughout this later work it has been 
 clear that increased turgor is present during abscission. In view of 
 the evidence given above, however, it would seem that turgor can 
 play only a secondary role, although the occurrence of increase in 
 turgor must not be ignored. 
 
 The bulging of the epidermis frequently noted as accompanying 
 abscission is evidence of increased internal pressure. In the pith the 
 cells next to those which are separating are in a collapsed condition 
 due to the pressure of the expanding separating cells. By various 
 experiments it can be shown that humid conditions favor and severe 
 drought prevents abscission. Richter and others have shown that 
 narcotic vapors which cause abscission also cause increased turgor by 
 increasing the proportion of sugar in starch-containing cells. This 
 increase in cell turgor becomes so great as to cause complete macera- 
 tion in certain types of tissues. The frequent presence of starch 
 grains in the separation layer of Nicotiana, part of which are prob- 
 ably converted into sugar as a result of subjection to illuminating 
 gas, indicates that there is probably an increase of turgor during 
 abscission, at any rate when induced by illuminating gas. 
 
 On the other hand, a more extensive examination of abscission in 
 certain plants indicates that all evidences of increased turgor may at 
 
380 University of California Publications in Botany [VOL. 5 
 
 times be absent. Such cases might be explained by the absence of 
 any considerable amount of starch in the cells concerned. Indeed, 
 the starch grains usually noted in the separation layer can not at 
 times be observed. This might also explain the fact that the bulging 
 of the epidermis and collapse of cells in the pith usually accompany- 
 ing abscission are sometimes absent. Also, starch grains are rarely 
 observed in the separation cells of Lycopersicum and Datura and in 
 these forms very little bulging of the epidermis occurs. Although 
 humid conditions favor abscission and drought prevents the process, 
 it has also been observed that drought has to be very severe before it 
 produces such a result. Other evidences for increased turgor derived 
 from the turgid appearance of the cells are mostly obtained after 
 abscission has started and, granting that the cells are isolated by dis- 
 solution of the middle lamella, more or less expansion due to release 
 of pressure is to be expected. 
 
 A critical examination of the separation cells during abscission 
 brings out several facts, other than those mentioned in the above 
 paragraph, which of themselves render inadmissible the theory that 
 cell separation is brought about by increased turgor. These are 
 as follows: 1. There seems to be no perceptible change in cell 
 shape or size during separation. 2. The increase in size of the inter- 
 cellular spaces does not necessarily take place first between the walls 
 at the " corners" of the cells, but may appear first as a longitudinal 
 streak between the lateral walls of the cells (pi. 51). 3. Cell isola- 
 tion may be incomplete in large numbers of cells still remaining 
 attached to each other. 4. Cell separation first becomes complete in 
 a narrow plane between only two tiers of cells before spreading later 
 to a larger number of cells. 5. The spreading of cell separation 
 itself is obviously hard to explain on the basis of the turgor theory. 
 
 In view "of the facts brought out in this discussion and the positive 
 evidence for the dissolution of the primary membranes, it should be 
 clear that increase in turgor, at least in the Solanaceae, is not the 
 direct cause of cell separation. Undoubtedly there is often great 
 increase in turgor during abscission, especially in certain types of 
 cells, but this increase, instead of being the direct initiating factor, 
 probably serves merely to hasten and facilitate the process. 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanacette 381 
 
 e. EXPERIMENTS ON THE AMOUNT OF SUGAR IN THE STEM 
 AND PEDICEL OF NICOTIAN A DURING ABSCISSION 
 
 After several experiments, all of which indicated the results ob- 
 tained below, the following experiment was performed. Experiment 
 2a was devised to show the change in the amount of sugar which 
 occurs in the tissues of the pedicel during abscission. Experiment 2& 
 was intended to show this same difference in a restricted region of the 
 stem just proximal to the separation layer. 
 
 Experiment 2a. Lot A included 200 pedicels of flowers which had 
 fallen a few minutes before being collected as a result of being sub- 
 jected to illuminating gas. Lot B included 200 pedicels of flowers 
 picked at the same time as those making up Lot A, but in which no 
 abscission was induced. The water extracts made with 10 cc. from 
 equal weights of the two lots were tested with surplus Fehlings solu- 
 tion. The precipitates formed upon boiling weighed : 
 
 A 68 mg. 
 
 B 95 mg. 
 
 Experiment 21}. This experiment was carried out in the same man- 
 ner as experiment 2a, but the precipitates in this case were of such 
 small quantity that no attempt was made to get actual figures as to 
 their weights. It was clear, however, merely from an examination of 
 the filter paper, that there was more precipitate in B than in A just 
 the reverse of Experiment 2a. The difference was evidently not as 
 great as in the latter experiment. 
 
 Experiment 2a seems to indicate that during abscission there is a 
 reduction of nearly one-third the normal amount of sugar in the 
 pedicel. Other preliminary experiments performed as abscission was 
 starting showed only a slight reduction in the amount of sugar in the 
 pedicel. Thus possibly the withdrawal of sugar commences with the 
 start of abscission. Experiment 2& indicates that there is probably a 
 slight increase in the amount of sugar in the limited region proximal 
 to the separation during abscission. It is possible that most of the 
 withdrawn sugar is used as a source for the energy required in the 
 active process of cell separation. The slight increase proximal to the 
 separation layer also shows that there is probably an increase in cell 
 turgor in the actual tissues which contain the separation layer, due 
 to the conversion of starch into sugar. 
 
382 University of California Publications in Botany [VOL. 5 
 
 /. POSSIBLE AGENCY ACTIVE IN THE DISSOLUTION OF 
 THE MIDDLE LAMELLA 
 
 The pectose of the middle lamella may be broken down into the 
 soluble pectin in three different ways by the action of an acid, 
 of an alkali, or of the enzyme pectosinase. Since it is doubtful 
 whether alkaline reactions in living cells frequently get strong enough 
 to affect the middle lamella, the probable active agency is limited 
 to the acid or the enzyme action. Up to the last few years very 
 little has been known about the action of enzymes concerned in 
 pectic digestion. It has been natural, therefore, for investigators 
 (cf. Wiesner, 1905, and Kubart, 1906) to consider the acid as prob- 
 ably the active agency. In this connection, it is well to state that I 
 have obtained distinct acid reactions with litmus from the base of the 
 corolla of Nicotiana during abscission. This would confirm Kubart, 
 who, it will be remembered, obtained similar reactions from the corolla 
 of Nicotiana. But in this case I sometimes obtained acid reactions 
 from the corolla when in the normal condition. Since these observa- 
 tions offer no detailed evidence that acidity has increased during 
 abscission to a degree higher than normal, their significance can well 
 be doubted. 
 
 The tissues of Datura give a distinct acid reaction to litmus in the 
 normal condition. Experiment 3 below shows a slight increase in 
 acidity during abscission. No acid reactions of much intensity are 
 given by the base of the pedicel of Nicotiana either in the normal or 
 abscissed condition. 
 
 Experiment 3. Lot A contained the bases of three pedicels cut 
 while abscission was going on. Lot B contained an equal weight 
 (6 gin.) of the bases of three pedicels cut in the normal condition. 
 These were extracted with water and the extracts made up to 10 cc. 
 each. By titration with 10 per cent NaOH and phenolphtalein the 
 following results were obtained : 
 
 A 0.75 cc. required to neutralize 
 
 B 0.6 cc. required to neutralize 
 
 A similar experiment on Nicotiana showed, however, that the nor- 
 mally low acidity of this genus is slightly reduced during abscission, 
 as indicated by the following results: 
 
 A 0.25 cc. required to neutralize 
 
 B 0.37 cc. required to neutralize 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceac 383 
 
 The normal acidity in Datura is high, but it is doubtful whether 
 the increase is large enough to account for the dissolution of the 
 middle lamella. At any rate, it is certain that acidity does not enter 
 into the problem in the pedicel of Nicotiana. We must, therefore, 
 fall back upon the enzyme action as probably responsible for the 
 process of cell separation. 
 
 Most hydrolysing processes characteristic of living cells aFe now 
 supposed to be due to the action of enzymes of different kinds. It 
 has been definitely claimed (cf. Atkins, 1916) that an enzyme which 
 has been called pectosinase is capable of breaking down the pectose 
 of which the middle lamella is composed. Add to this the fact that the 
 action of enzymes has been shown, as has also the process of abscission, 
 to be very sensitive to all kinds of changes in the external environment, 
 and it is fairly safe to assume that the method of cell separation is 
 fundamentally an enzyme problem. Irrefutable proof of this could 
 be obtained only by testing for the activity of pectosinase during the 
 early stages of abscission and by demonstrating the absence or in- 
 activity of this enz3 r me in species where abscission does not occur. 
 
 ABSCISSION OF THE STYLE AND COROLLA 
 
 Abscission of the corolla in Nicotiana was described by Kubart 
 and it may be said at once that the observations herein described 
 
 agree entirely with his. Abscission of the 
 corolla is brought about by the separation, 
 without any previous cell divisions or 
 elongations, of living cells at the base of 
 the corolla tube. The separation layer, 
 which is in no way morphologically differ- 
 entiated from the neighboring tissue, is 
 located about 1 mm. from the point of in- 
 sertion of the corolla on the receptacle. It 
 thus occurs in the distal part of a region 
 where intergradation of cell shape, between 
 the isodiametric cells of the receptacle 
 and the more or less elongated cells of the 
 corolla, is apparent. The separation cells 
 which are in this region of intergradation 
 
 Fig. 6. Longitudinal radial . . 
 
 section of the base of the are not isodiametric but are more or less 
 
 corolla tube of Nicotiana, elongated parallel to the long axis of the 
 showing the method of ab- 
 scission, corolla. All the cells in cross-section of the 
 
384 
 
 University of California Publications in Botany [VOL. 5 
 
 base of the corolla tube at about the level of the separation layer seem 
 to be involved in the process except the epidermal cells and the 
 tracheae. The process of cell isolation in this case may spread up and 
 down for quite a distance between the epidermis and tracheae, thus 
 involving a large number of cells (fig. 6). 
 
 Abscission of the corolla in Datura differs slightly from that in 
 Nicotiana,. As in the latter, there is no differentiated separation 
 layer, separation occurring in cells which are not visibly different 
 from other cells of the corolla. Cells more or less elongated are in- 
 volved, as in Nicotiana, but in Datura the region of separating cells 
 is limited to certain tissues that is to say, not all the cells across 
 the base of the corolla tube at about the level of the separation layer 
 are involved in the process of abscission. The base of the corolla in 
 Datura is characterized by distinct longitudinal ridges which alternate 
 with deep grooves. Thus, a cross-section of a portion of the base of 
 the corolla appears as in fig. 8. Cell separation fails to occur in the 
 outside ridges at the level of the separation layer, so that, looking at 
 the base of the corolla tube from the outside during abscission, one 
 sees separate crescent-shaped regions of macerating cells alternating 
 with cells which are not separating (fig. 7). This is explained when 
 a cross-section is taken (fig. 8), which shows that several vascular 
 bundles, the cells of which do not sep- 
 arate, are collected in the outside ridges. 
 
 Abscission of the style occurs nor- 
 mally in Nicotiana and Datura a short 
 time before the corolla has fallen. So 
 far as it was possible to determine, the 
 process of abscission is exactly the same 
 in the style as in the corolla. A separa- 
 tion of very small, more or less elongated 
 cells takes place at the base of the style 
 without any external indication such as Fig. 7 
 
 frequently occurs in the pedicel of the c region of separating cells. 
 
 'a c 
 Fig. 8 
 
 a vascular bundle. 
 c region of cell separation. 
 5 region of no cell separation, 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 385 
 
 flower. As in the case of the corolla, there is no structure which might 
 be interpreted as a differentiated separation layer. Separation occurs 
 in a region of intergradation in cell shape between the spherical cells 
 of the ovary and the cells of the style, which are elongated parallel 
 with the long axis of the style. 
 
 TIME OF ABSCISSION 
 1. REACTION TIME 
 
 a. EEACTION TIME IN NOKMAL ABSCISSION 
 
 The term "reaction time" is used in referring to two rather dis- 
 tinct subjects. First, we may have a reaction time represented by 
 the time intervening between anthesis and normal abscission due to 
 lack of fertilization. Second, we may have a reaction time which has 
 to do with the period between the application of the stimulus and 
 flower-fall in ' ' spontaneous ' ' abscission. The reaction times in normal 
 abscission were discussed in an earlier communication in the case of 
 two F! species hybrids of Nicotiana and their parents. The state- 
 ments there made have been repeatedly verified and in addition a 
 considerable amount of data has been accumulated in regard to the 
 time of abscission in other species of Nicotiana and in the genus 
 Lycopersicum. In the case of the former the observations were also 
 made upon the abscission of the corolla, the effect of pollination on 
 reaction time, and the reaction time in "spontaneous" abscission. 
 
 In determining the abscission times for the hybrid F t H154 (cf. 
 page 386) , a great variation was noted in the normal reaction time. In 
 the case of the hybrid Fj H179 (cf. page 386), however, and in other 
 species or varieties investigated, very little variation in the time of 
 abscission has been noted. The range of variation in these species 
 and varieties practically always falls within two or three days and a 
 large number of observations gives identical times, as far as the number 
 of days is concerned, in the case of seven to ten flowers. There is a cer- 
 tain variation in the length of the reaction times in different flowers on 
 the same plant, but the plants of a species do not differ from one an- 
 other in their average reaction times. It was noticed that the figures 
 were approximately the same whether the averages were based on the 
 records of four or five flowers or of a considerably larger number ; 
 thus the results given in the following table may be considered con- 
 clusive. Where the number of flowers involved is less than four, 
 
386 
 
 University of California Publications in Botany {.VOL. 5 
 
 however, the results serve merely as approximate estimate of the 
 abscission reaction time. 
 
 In obtaining the records tabulated below a separate tag was sup- 
 plied for each flower. This tag was put on the flower at the begin- 
 ning of the observation, the plant visited twice a day thereafter and 
 records kept on the tags, which were left on the flower until the close 
 of the observation. If a flower fell upon being tapped or shaken, 
 abscission was considered to have occurred and the date was recorded 
 on the tag, which was then collected. Similarly, in the case of the 
 records for abscission of the corolla, a slight pull had to be applied 
 before it could be determined whether abscission had occurred. As a 
 means of preventing fertilization, the, stigma was cut away in addi- 
 
 TABLE 1 
 
 
 I 
 
 II 
 
 III 
 
 IV 
 
 V 
 
 
 Time from 
 
 Time from 
 
 Time from 
 
 Time from 
 
 Time from 
 
 
 bud 1 to 
 anthesis 
 
 pollination to 
 mature fruit 
 
 pollination to 
 abscission of 
 
 anthesis to 
 abscission of 
 
 anthesis to 
 normal flower 3 
 
 Designation of 
 species or variety 
 
 
 
 corolla 
 
 corolla, 
 unpollinated 
 
 fall 
 
 
 No. 
 flowers 
 
 Avg. 
 No. 
 days 
 
 No. 
 
 flowers 
 
 Avg. 
 No. 
 days 
 
 No. 
 flowers 
 
 Avg. 
 
 No. 
 days 
 
 No. 
 flowers 
 
 Avg. 
 No. 
 
 days 
 
 No. 
 flowers 
 
 Avg. 
 No. 
 days 
 
 Fi H38 
 
 
 
 
 
 
 
 
 
 10 
 
 18 
 
 F t H179 
 
 4 
 
 8 
 
 4 
 
 14 
 
 15 
 
 3 
 
 23 
 
 8 2 
 
 6 
 6 
 
 20 
 
 7 
 
 F t H36 
 
 
 
 
 
 7 
 
 4 
 
 3 
 
 6 
 
 
 
 N. sylvestris 
 
 
 
 10 
 
 11 
 
 6 
 
 4 
 
 9 
 
 4 2 
 
 6 
 6 
 
 9 
 
 15 
 
 N. Tabacum 
 
 
 
 3 
 
 12 
 
 2 
 
 4 
 
 2 
 
 6 
 
 3 
 
 5 
 
 "Maryland" 
 
 
 
 
 
 
 
 
 
 
 
 N. Bigelovii 
 
 
 
 5 
 
 10 
 
 5 
 
 3 
 
 5 
 
 5 
 
 8 
 
 17 
 
 var. Wallace! 
 
 
 
 
 
 
 
 
 
 
 
 N. Bigelovii 
 
 
 
 5 
 
 19 
 
 5 
 
 2 
 
 6 
 
 5 
 
 G 
 
 no fall 
 
 "Porno" 
 
 
 
 
 
 
 
 
 
 
 
 N. Bigelovii f 
 
 
 
 
 
 
 
 
 
 
 
 16 
 
 var. typica \ 
 
 
 
 
 
 
 
 
 
 3 
 
 no fall 
 
 N. Bigelovii 
 
 
 
 4 
 
 21 
 
 1 
 
 4 
 
 2 
 
 8 
 
 
 
 (hybrid?) 
 
 
 
 
 
 
 
 
 
 
 
 N. multivalvis 
 
 
 
 3 
 
 18 
 
 2 
 
 4 
 
 
 
 2 
 
 no fall 
 
 N. quadrivalvis 
 
 
 
 5 
 
 20 
 
 5 
 
 7 
 
 5 
 
 2 
 
 6 
 
 no fall 
 
 N. suaveolens 
 
 
 
 
 
 8 
 
 8 
 
 8 
 
 8 
 
 7 
 
 13 
 
 N. Sanderae 
 
 
 
 15 
 
 15 
 
 5 
 
 4 
 
 5 
 
 6 
 
 6 
 
 9 
 
 N. rustica var. 
 
 
 
 
 
 
 
 3 
 
 4 
 
 4 
 
 6 
 
 brasilia 
 
 
 
 
 
 
 
 
 
 
 
 N. rustica 
 
 
 
 7 
 
 6 
 
 7 
 
 3 
 
 4 
 
 4 
 
 
 
 (Winnebago) 
 
 
 
 
 
 
 
 
 
 
 
 N. rustica var.? 
 
 
 
 4 
 
 15 
 
 4 
 
 2 
 
 5 
 
 3 
 
 6 
 
 5 
 
 Lycopersicum 
 esculentum 
 
 4 
 
 6 
 
 
 
 4 
 
 3 
 
 4 
 
 8 
 
 17 
 
 9 
 
 1 The buds recorded here were of such size that the corolla and calyx were of the 
 same length. 
 
 2 Sterile pollen applied to the stigma. 
 
 8 By normal flower-fall is meant fall due to lack of fertilization. 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 387 
 
 tion to removing the anthers before anthesis. Various experiments 
 had shown that such an operation on the flower does not induce 
 abscission or affect its normal physiological condition to any great 
 extent. 
 
 F! H154 is Nicotiana Tabacum var. macrophylla (U. C. B. G. 
 22/07) X A 7 - sylvestris (U. C. B. G. 69/07). F H179 is N. Tabacum 
 "Cuba" (U. C. B. G. 200/14) X N. sylvestris. F t H36 i^N~rsylvestris 
 X N. Tabacum var. angustifolia (U. C. G. B. 68/07). 
 
 The results given in table 1 indicate, in the first place, that the 
 different species differ considerably in all the types of abscission 
 reaction times considered, and, in the second place, that on the aver- 
 age, application of a fertile pollen to the stigma tends to shorten 
 the time between anthesis and abscission of the corolla by two days. 
 The one apparent exception to this statement is Nicotiana suaveolens, 
 but in this case the pollinated flowers fell five or six days later 
 than the corolla, indicating that growth of the pollen had not pro- 
 ceeded very far. Records on F x H179 and N. sylvestris indicate that 
 sterile pollen does not have the same effect on abscission that fertile 
 pollen does. This would seem to show that here the effect of pollina- 
 tion upon the postfloration phenomena is not due, as Fitting (1909) 
 has found in orchids, to mere mechanical or .chemical stimulation of 
 the stigma by the pollen. This much being certain, the question still 
 remains whether the results obtained depend upon fertilization or are 
 due to the growth of the pollen tubes down through the style. 
 
 According to East (1915), working on self -sterility in Nicotiana 
 hybrids, the pollen tubes reach the ovary, in cases of cross-pollination, 
 three or four days after application of pollen to the stigma. Since in 
 all cases recorded above cross-pollination was carried on and since in 
 most cases the corolla was not thrown off until three or four days 
 after application of pollen to the stigma, it is possible that fertiliza- 
 tion is the important factor in shortening the time between anthesis 
 and abscission of the corolla. In N. qttadrivalvis, however, the corolla 
 was thrown off within eighteen hours after pollination, whereas, when 
 pollination is prevented, the corolla may remain on the flower for 
 fifty-seven hours. If East's conclusions are correct, this would seem 
 to indicate that the shortening of the reaction time in abscission of 
 the corolla is due to some stimulation of the style by the pollen tubes 
 and not to fertilization. This conclusion, however, could be doubted 
 even here, because the style of N. quadrivalvis is very short, so that 
 the pollen tubes might reach the ovary in a much shorter time than is 
 
388 University of Californw Publications in Botany [Vol.. ~> 
 
 required in larger flowers. An attempt was made to get further data 
 on this point by removing the style several Jiours after application of 
 pollen before the pollen tubes could possibly have reached the ovary. 
 This operation occasionally causes the whole flower to fall, and since 
 in such cases abscission in the pedicel occurs before fall of the corolla, 
 no results in regard to the latter organ are obtained. The possible 
 effect of the operation on the abscission of the corolla was checked 
 by control tests of unpollinated flowers in which the styles had also 
 been removed. This can also be checked by a comparison with the 
 periods of. time given in table 1, column III. 
 
 It was found in three flowers of t H179 that, when the style was 
 removed three days after pollination, the corolla was, on the average, 
 thrown off three days after anthesis. The control test for this experi- 
 ment gave in three flowers an average of five days. Where the style' 
 was removed two days after anthesis, four flowers gave an average of 
 three days. Where the style was removed one day after pollination, 
 the corolla was abscissed in five flowers an average of three days after 
 anthesis. A control test gave in this case an average of five days for 
 five flowers. Finally, the style was removed in seven flowers seventeen 
 hours after pollination. The seven flowers gave in this case an average 
 of four days for the time between anthesis and fall of the corolla. A 
 control test for this last case gave for five flowers an average of five 
 days. 
 
 These experiments were repeated with N. sylvestris. In one case 
 where the style was removed in three flowers two days after pollina- 
 tion, the corolla was thrown off on an average of four days after 
 anthesis. A control test of this case gave an average of six days for 
 three flowers. In another case the style was removed in three flowers 
 one day after pollination. In this case the corolla was abscissed 011 
 an average of three days after anthesis. 
 
 The results given in the above paragraphs indicate definitely that 
 it is the stimulation of stylar tissues caused by the growth of the 
 pollen tubes which shortens the time between anthesis and abscission 
 of the corolla. They also show that the removal of the style has no 
 appreciable effect on the abscission of the corolla. It is evident from 
 the results given that the influence of the pollen is seen as early as 
 seventeen hours after pollination, and it is possible that the effect may 
 be manifested even earlier. It is significant that the period given in 
 the case where the style was removed seventeen hours after pollina- 
 tion is one day longer than in the case where the style was removed 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 389 
 
 twenty-four hours after pollination. This may possibly indicate that 
 in the first case the influence of the pollen tubes has diminished, be- 
 cause of the shortening of the period which they have had for growth. 
 If this is the case, it is reasonable to suppose that the influence of the 
 growing pollen tube increases up to twenty-four hours after pollina- 
 tion as the pollen tube lengthens. Thus, at six hours after pollination 
 it is possible that no effect of the pollen tubes would be~ noticeable, 
 while twenty-four hours after pollination the entire influence of the 
 growing pollen tube has been exerted. 
 
 The effect of pollination on the time between anthesis and flower- 
 fall was tested by experiments similar to those described above. 
 Results in such experiments are difficult to obtain because removal of 
 the style frequently causes the premature fall of the flower. If the 
 flower fell before abscission of the corolla, the fall was considered 
 premature, as the result of the removal of the style, and the record of 
 that particular flower not considered. Since under ordinary condi- 
 tions pollinated flowers remain on the plant, it is to be expected that 
 the stimulation of the stylar tissues by the pollen tubes, if it has any 
 influence at all, would increase the length of time between anthesis 
 and flower-fall. Granting the truth of this assumption, any reduction 
 in time between anthesis and fall can be considered as the result of 
 removal of the style. 
 
 In one test on ten flowers of F x H179, where the style was removed 
 two days after pollination, flower-fall occurred on an average of seven 
 days after anthesis. A control test in this case also gave seven days 
 for ten flowers. This time is approximately the same (the actual 
 average calculated to the tenth of a day was 6.7) as those given in 
 table 1, column V, for the time between anthesis and normal flower- 
 fall due to lack of fertilization. A similar test on six flowers of N. syl- 
 veslris, where the style was removed two days after pollination, gave 
 an average of thirteen days. The time for this species in table 1, 
 column V, is fifteen days. > 
 
 These two records indicate that the stimulation of the stylar tissues 
 by the growing pollen tubes has no effect on the time between anthesis 
 and flower-fall. In the second case above, and also perhaps in the 
 first, the stimulation of the style seems to have shortened the time 
 somewhat, but in this case the result can be explained by the effect 
 of the later removal of the style. 
 
390 rniuersity of California Publications in Botani/ [VOL. 5 
 
 b. REACTION TIME IN "SPONTANEOUS" ABSCISSION 
 
 Exact data in regard to the reaction time can be given only in 
 two definite cases. The observations in these cases were made on 
 small shoots of the plant to be considered, which were placed in water 
 and inserted under a bell-jar containing 1.5 per cent illuminating 
 gas. After several hours, the material was shaken 'every fifteen min- 
 utes to determine when the first flower fell. F x H179 and N. Tabacum 
 1 ' Maryland " were selected as material for the experiments because 
 these forms were found most sensitive and thus react regularly and 
 quickly to stimuli. Abscission occurs in the pedicel of F x H179 seven 
 hours after insertion into 1.5 per cent illuminating gas at a tempera- 
 ture of approximately 19 C. The smaller buds begin to fall first, 
 but are followed in a short time by the open flowers. Abscission 
 occurs in X. Tabacum " Mary land" in eight hours under the above 
 conditions. 
 
 The remainder of the data having to do with the reaction time in 
 spontaneous abscission is in the form of approximate estimates derived 
 from the results of experiments on the induction of abscission. In 
 the case of abscission induced by illuminating gas most species which 
 shed their flowers in 1.5 per cent illuminating gas do so after ten or 
 fifteen hours at room temperature. 
 
 There remains now to be considered the reaction time in cases of 
 flower-fall due to mechanical injury. The results along this line are 
 largely derived from tables 2, 3, 4, and 5, which, however, were 
 arranged to show more particularly the comparative effect of different 
 types of injury, as causing or not causing abscission in flowers of 
 various ages. These tables might as well be presented under the 
 heading "Experimental Induction of Abscission by Mechanical In- 
 jury" (page 405), but since it is necessary to draw certain conclusions 
 from them in regard to the time of abscission they are presented and 
 explained at this time. 
 
 Tables 2, 3, 4, and 5, which follow, serve to record the results of 
 a number and variety of experiments all designed to show the relation 
 of mechanical injury to abscission. It was very soon discovered while 
 carrying on the experiments that the effect of injury depends to a 
 large extent upon the age of the flower. Now the age of the flower 
 can be most conveniently measured by determining the increase in 
 size of growing parts such as the corolla and ovary. Thus it was 
 necessary in each case to record the size of the flower size being a 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 391 
 
 criterion of age upon which the test was being made. This was done 
 by noting on the tag which was supplied for each flower (cf. page 385) 
 the length of the corolla in millimeters, the condition of the corolla, 
 or any other condition of the flower which would serve to indicate its 
 age. The period of development of the flower and fruit is divided 
 into several arbitrary stages, each of which is designated by a Roman 
 numeral in the second column of the tables. Where the~mimber of 
 flowers designated in the first column are nearly in the same stage 
 of development only one numeral appears in the table, but where the 
 range in size of the flowers is quite extensive two numerals appear, 
 representing the range in size within which the flowers were found at 
 the time of the experiment. The stages of floral development which 
 each Roman numeral represents are given below. 
 
 Bud 
 
 I corolla 2 mm. to 5 mm. in length 
 
 II corolla 6 mm. to 10 mm. in length 
 
 III corolla 11 mm. to 15 mm. in length 
 
 IV corolla 16 mm. to 20 mm. in length 
 
 V corolla 21 mm. to 30 mm. in length 
 
 VI corolla 31 mm. to 40 mm. in length 
 
 VII corolla 41 mm. to 50 mm. in length 
 
 Flower 
 
 VIII corolla opening 
 
 IX anthesis 
 
 X 2 days after anthesis 
 
 XI corolla withering 
 
 Fruit 
 
 Immature 
 
 XII fruit 5 mm. to 8 mm. in length 
 
 XIII fruit 9 mm. to 10 mm. in length 
 
 Mature 
 XIV fruit 11 mm. to 12 mm. in length 
 
 The operation of injuring the flower consisted largely in removing, 
 by cutting away with a sharp safety razor blade, entire floral organs 
 or parts of them. In some cases, however, organs were only slit longi- 
 tudinally with a sharp knife or merely punctured with the point of a 
 pair of forceps. 
 
 Several types of injury that remove the style, stigma or stamens 
 before pollination may -cause fall by preventing fertilization. It is 
 evident, therefore, that fall occurring after such an operation per- 
 formed on the flower before anthesis may be due to lack of fertiliza- 
 tion and not to the injury. If, however, the fall occurs within the 
 minimum time elapsing between anthesis and normal flower-fall due 
 
392 
 
 University of California Publications in Botany [VOL. 5 
 
 to lack of fertilization, it can be safely concluded that the fall is due 
 to the effect of the injury. This minimum time is about seven days 
 for N. Langsdorffii. It can be safely said, therefore, that any fall 
 occurring in less than seven days after injury to the flower near 
 anthesis is due directly to the effect of the injury. In cases where the 
 stamens or style are removed in flowers younger than those at anthesis, 
 
 TABLE 2 
 
 EFFECT OF DIFFERENT TYPES OF INJURY IN CAUSING FLOWER FALL IN 
 N. Langsdorffii var. ffrandi flora 
 
 No. 
 flowers 
 
 Size or 
 condition 
 of flowers 
 
 Injury to 
 
 Avg. No. 
 days before 
 fall of 
 remaining 
 organs 
 
 Calyx 
 
 Corolla 
 
 Stamens 
 
 Pistil 
 
 Pedicel 
 
 
 '10 
 
 II-VII 
 
 all cut 
 
 all cut 
 
 all cut 
 
 all cut 
 
 
 1 
 
 
 10 
 
 VII-XI 
 
 
 < i 
 
 t < 
 
 t 
 
 
 2 
 
 H - 
 
 10 
 
 XII-XIII 
 
 
 a 
 
 < t 
 
 t 
 
 
 3 
 
 
 . 2 
 
 XIV 
 
 
 1 1 
 
 1 1 
 
 ' 
 
 no fall 
 
 
 ' 4 
 
 I-II 
 
 I cut 
 
 1 cut 
 
 1 1 
 
 style cut 
 
 7 
 
 
 3 
 
 III-VII 
 
 
 ( < 
 
 1 1 
 
 ' 
 
 7 
 
 
 3 
 
 VIII-IX 
 
 
 tl 
 
 1 1 
 
 i 
 
 
 7 
 
 }, 
 
 5 
 
 XI -XII 
 
 
 It 
 
 1 1 
 
 
 
 
 no fall 
 
 u - 
 
 2 
 
 XI-XII 
 
 
 11 
 
 < t 
 
 style and 
 
 
 5 
 
 
 
 
 
 
 
 part of 
 
 
 
 
 
 
 
 
 
 ovary cut 
 
 
 
 . 3 
 
 XIII-XIV 
 
 1 1 
 
 I i 
 
 it 
 
 1 1 
 
 
 4 
 
 f 5 
 
 V-VII 
 
 
 i cut 
 
 1 1 
 
 I style cut 
 
 
 7 
 
 oJ 5 
 
 VIII-IX 
 
 
 < 
 
 it 
 
 1 1 
 
 
 6 
 
 2 
 
 XI 
 
 
 < I 
 
 i 
 
 tt 
 
 
 9 
 
 I 3 
 
 XII 
 
 
 t i 
 
 " 
 
 1 1 
 
 
 no fall 
 
 ( 4 
 
 I 
 
 all cut 
 
 
 
 
 
 3 
 
 &\ I 
 
 II 
 
 t i 
 
 
 
 
 
 no fall 
 
 112 
 
 III-XII 
 
 it 
 
 
 
 
 
 1 1 
 
 f 2 
 
 V-VII 
 
 slit on sides 
 
 2 slit on 
 
 
 
 
 7 
 
 
 
 
 to base 
 
 2 sides 
 
 
 
 
 
 
 
 
 
 to base 
 
 
 
 
 
 1 8 
 
 m-xi 
 
 
 
 
 
 
 no fall 
 
 t? " 
 
 3 
 
 II 
 
 
 
 
 ovarv slit 
 
 
 3 
 
 
 3 
 
 V-VII 
 
 
 
 
 1 1 
 
 
 5 
 
 
 3 
 
 IX 
 
 
 
 
 1 1 
 
 
 2 
 
 
 k 2 
 
 XII 
 
 
 
 
 1 1 
 
 
 5 
 
 
 ' 5 
 
 IX 
 
 
 
 anthers 
 
 
 
 10 
 
 
 
 
 
 
 
 all cut 
 
 
 
 
 \ 2 
 
 V-VII 
 
 
 
 all cut 
 
 
 
 7 
 
 I 3 
 
 VII -X 
 
 
 
 1 1 
 
 
 
 9 
 
 
 4 
 
 VIII 
 
 
 
 
 stigma cut 
 
 
 9 
 
 g 
 
 2 
 
 V-VII 
 
 
 
 
 style cut 
 
 
 7 
 
 
 2 
 
 VII-X 
 
 
 
 
 i i 
 
 
 10 
 
 
 f 4 
 
 XIV 
 
 all cut 
 
 
 
 all cut 
 
 
 no fall 
 
 
 2 
 
 XIII 
 
 i cut 
 
 
 
 i cut 
 
 
 2 
 
 
 
 2 
 
 XIV 
 
 n 
 
 
 
 1 1 
 
 
 no fall 
 
 
 4 
 
 XIII-XIV 
 
 slit to base 
 
 
 
 slit to base 
 
 
 
 
 
 f 10 
 
 II 
 
 
 
 
 
 slit to base 
 
 a 
 
 
 3 
 
 IX 
 
 
 
 
 
 1 1 
 
 n 
 
 
 6 
 
 I-VIII 
 
 
 
 
 
 1 cut 
 
 1 1 
 
 1 H 
 
 
 
 
 
 
 
 i through 
 
 
 
 8 
 
 II-XII 
 
 
 
 
 
 2 cuts 
 
 1 1 
 
 
 
 
 
 
 
 
 
 i through 
 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 393 
 
 allowance must be made for the approximate number of days preced- 
 ing anthesis. Thus, if a flower of the above species is injured three 
 days before anthesis, the fall can not be assigned to the injury unless 
 it occurs before ten daj^s have elapsed. The minimum time for 
 Fj H179 is about five days ; thus, any time of five days or more recorded 
 on a flower, injured near anthesis, was considered as "no fall." The 
 minimum time for Lycopersicum is about six days. 
 
 Finally, it is necessary to state that the process of reaction to the 
 different types of injury recorded in the following tables was by no 
 means impeded by low temperatures. Nicotiana Langsdorffii was 
 tested out in a greenhouse where the average temperature approxi- 
 mated 75 F. The tests on F x H179 and Lycopersicum were per- 
 formed in the botanical garden of the University during July and 
 August, when the temperature was also comparatively high. 
 
 The following statement of results is derived in great part but not 
 entirely from the foregoing tables. It has been noticed that cutting 
 off the freshly opened flower at the tip of the pedicel causes the 
 remainder of the pedicel to be thrown off in from ten to fifteen hours, 
 but after the same operation on developed capsules the pedicel re- 
 mains firm from thirty-six to ninety-six hours after the injury. 
 Removal of the calyx causes the fall of buds in two or three days, 
 depending upon the age of the bud. Removal of half the calyx 
 together with two-thirds of the corolla and all the stamens causes 
 fall in one to four days, depending upon the age of the flower. A 
 
 TABLE 3 
 
 EFFECT OF POLLINATION OF FLOWERS OF -N. Langsdorffii var. grandiflora ON 
 REACTION TO INJURY 
 
 No. 
 flowers 
 
 Pollination 
 
 Injury 
 
 Avg. No. 
 days before 
 fall 
 
 41 
 
 pollinated when injured 
 not pollinated 
 
 calyx and 
 
 stamens cut 
 
 no fall 
 10 
 
 , / 4 
 
 pollinated when injured 
 
 calyx " 
 
 i corolla cut 
 
 no fall 
 
 1 5 
 
 not pollinated 
 
 
 tt 
 
 8 
 
 
 No. days after pollination when 
 
 
 
 
 
 injured 
 
 
 
 
 f 2 
 
 1 
 
 all organs 
 
 cut at tip of pedicel 
 
 2 
 
 1 2 
 
 2-6 
 
 
 tt 
 
 2 
 
 c j 2 
 
 7-8 
 
 
 < f 
 
 2 
 
 1 3 
 
 2 
 
 4 calyx, 
 
 corolla, stamens. 
 
 4 
 
 I 
 
 
 style cut 
 
 
 f 3 
 
 4-5 
 
 
 11 
 
 5 
 
 dJ 2 
 
 6-7 
 
 
 ( i 
 
 2 
 
 d l 1 
 
 9 
 
 
 
 
 3 
 
 I 1 
 
 9 
 
 
 i ( 
 
 no fall 
 
394 
 
 University of California Publications in Botany [ V L. 5 
 
 transverse cut through the entire flower which passes through the 
 middle of the ovary causes fall in one to two days. A similar oper- 
 ation in the case of maturing fruits changes the date of fall to 
 four to eight days. Kemoval of half the corolla and all the stamens 
 causes fall of buds in one day and the fall of young flowers in two to 
 three days. Removal of the stamens or style in buds causes fall in 
 
 TABLE 4 
 EFFECT OF DIFFERENT TYPES OF INJURY IN CAUSING FLOWER FALL IN FJI179 
 
 No. 
 
 flowers 
 
 Size or 
 condition 
 of flowers 
 
 Injury to 
 
 Avg. No. 
 days before 
 fall of 
 remaining 
 organs 
 
 Calyx 
 
 Corolla 
 
 Stamens 
 
 Pistil 
 
 Pedicel 
 
 / 9 
 
 11- VIII 
 
 i cut 
 
 cut 
 
 all cut 
 
 style cut 
 
 
 1 
 
 a \ 6 
 
 XI-XV 
 
 < t 
 
 i i 
 
 t t 
 
 
 
 
 no fall 
 
 {4 
 
 III-VII 
 
 
 i cut 
 
 1 1 
 
 
 
 1 
 
 10 
 
 VIII-IX 
 
 
 < i 
 
 11 
 
 
 
 no fall 
 
 c 10 
 
 v-vm 
 
 
 I I 
 
 
 
 
 " 
 
 r 9 
 
 I 
 
 all cut 
 
 
 
 
 
 2 
 
 
 7 
 
 II 
 
 
 
 
 
 
 2 
 
 
 3 
 
 II 
 
 
 
 
 
 
 no fall 
 
 
 4 
 
 III-IV 
 
 
 
 
 
 
 3 
 
 
 
 6 
 
 III-IV 
 
 
 
 
 
 
 no fall 
 
 
 1 
 
 V 
 
 
 
 
 
 
 2 
 
 
 4 
 
 V-VII 
 
 
 
 
 
 
 no fall 
 
 
 2 
 
 IX 
 
 
 
 
 
 
 " 
 
 
 ' 7 
 
 III-IV 
 
 
 
 tt 
 
 
 
 2 
 
 
 1 
 
 V 
 
 
 
 4< 
 
 
 
 5 
 
 e- 
 
 3 
 
 V 
 
 
 
 tt 
 
 
 
 no fall 
 
 
 6 
 
 VI -VII 
 
 
 
 a 
 
 
 
 < t 
 
 {5 
 
 II-VIII 
 
 
 
 
 " 
 
 
 2 
 
 4 
 
 VII -VIII 
 
 
 
 
 *< 
 
 
 no fall 
 
 
 i 
 
 II 
 
 1 slit on 2 
 
 1 slit on 2 
 
 
 
 
 5 
 
 
 
 
 sides to 
 
 sidos to 
 
 
 
 
 
 
 
 
 base 
 
 base 
 
 
 
 
 
 
 1 
 
 II 
 
 tt 
 
 " 
 
 
 
 
 no fall 
 
 
 9 
 
 I V-VII 
 
 a 
 
 (i 
 
 
 
 
 tt 
 
 > H 
 
 9 
 
 II 
 
 2 slits on 2 
 
 2 slits on 2 
 
 
 
 
 1 
 
 
 
 
 sides to 
 
 sides to 
 
 
 
 
 
 
 
 
 base 
 
 base 
 
 
 
 
 
 
 2 
 
 II-IV 
 
 st 
 
 11 
 
 
 
 
 4 
 
 
 5 
 
 V-VII 
 
 11 
 
 tt 
 
 
 
 
 no fall 
 
 
 5 
 
 II-V 
 
 punctured 
 
 punctured 
 
 
 ovary 
 
 
 2 
 
 
 
 
 on both 
 
 on both 
 
 
 punctured, 
 
 
 
 
 
 
 sides 
 
 sides 
 
 
 small hole 
 
 
 
 h- 
 
 3 
 
 VI -VII 
 
 " 
 
 " 
 
 
 < i 
 
 
 no fall 
 
 
 3 
 
 VII -XI 
 
 tt 
 
 t < 
 
 
 tt 
 
 
 2 
 
 
 3 
 
 Il-III 
 
 i i 
 
 tt 
 
 
 tt 
 
 
 2 
 
 
 3 
 
 VI-X 
 
 
 
 < i 
 
 
 " 
 
 
 2 
 
 
 15 
 
 III-XII 
 
 
 
 
 
 1 slit to 
 
 no full 
 
 
 
 
 
 
 
 
 base 
 
 
 i- 
 
 5 
 
 
 
 
 
 
 punctur'd 
 
 1 
 
 
 
 
 
 
 
 many 
 
 
 l 
 
 
 
 
 
 
 times 
 
 
 f 6 
 
 XIV 
 
 i cut 
 
 
 
 capsule 
 
 
 4 
 
 j J 
 
 
 
 
 
 i cut 
 
 
 
 I 3 
 
 XIV 
 
 " 
 
 
 
 < t 
 
 
 no fall 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 395 
 
 two to four days. Severe injury of any kind to the ovary causes fall 
 in one to two days. 
 
 The figures given above for the reaction time in cases of abscission 
 following mechanical injury, together with a more detailed considera- 
 tion of the tables, indicate that the reaction time, in general, does not 
 depend so much on the type of injury as on the age of the flower 
 concerned. What connection there is between the type of injury and 
 the reaction time seems to be based, except in cases of injury to the 
 ovary, on the relation of the amount of material removed to the 
 amount remaining. Thus, cutting off the flower at the tip of the 
 pedicel causes abscission of the remaining pedicel more quickly than 
 any other type of injury. One exception to this statement is seen, as 
 
 TABLE 5 
 
 EFFECT OF DIFFERENT TYPES OF INJURY IN CAUSING FLOWER FALL IN 
 Lycopersicum esculentum 
 
 No. 
 flowers 
 
 Size or 
 condition 
 of flowers 
 
 Injury to 
 
 Avg. No. 
 days before 
 fall of 
 remaining 
 organs 
 
 Calyx 
 
 Corolla 
 
 Stamens 
 
 Pistil 
 
 Pedicel 
 
 a- 
 
 ' 4 
 4 
 
 I 
 II -VIII 
 
 all cut 
 
 i I 
 
 
 
 
 
 no fall 
 
 a 
 
 
 . 6 
 
 XII 
 
 ( < 
 
 
 
 
 
 14 
 
 
 r 3 
 
 XII 
 
 
 
 
 entire 
 
 
 2 
 
 
 
 
 
 
 
 ovary cut 
 
 
 
 
 3 
 
 XII1-XIV 
 
 
 
 
 ovary 
 
 
 no fall 
 
 
 
 
 
 
 
 punctured 
 
 
 
 
 
 
 
 
 
 4 times on 
 
 
 
 !>- 
 
 
 
 
 
 
 top 
 
 
 
 
 1 
 
 XII 
 
 
 
 
 < < 
 
 
 3 
 
 
 4 
 
 XII 
 
 
 
 
 ovary 
 
 
 2 
 
 
 
 
 
 
 
 punctured 
 
 
 
 
 
 
 
 
 
 4 times on 
 
 
 
 
 
 
 
 
 
 side 
 
 
 
 
 , 3 
 
 XIV 
 
 
 
 
 it 
 
 
 no fall 
 
 
 ' 4 
 
 II 
 
 punctured 
 
 punctured 
 
 
 ovary 
 
 
 9 
 
 
 . 
 
 
 at base 
 
 
 
 punctured 
 
 
 
 c- 
 
 
 
 
 
 
 once on 
 
 
 
 
 
 
 
 
 
 side 
 
 
 
 
 L 4 
 
 VIII 
 
 (i 
 
 1 1 
 
 
 1 1 
 
 
 4 
 
 
 r 4 
 
 II- VIII 
 
 i cut 
 
 i cut 
 
 
 
 
 no fall 
 
 
 4 
 
 VIII -IX 
 
 < < 
 
 < 4 
 
 
 tt 
 
 
 5 
 
 d 
 
 3 
 
 I-II 
 
 
 i < 
 
 
 ovary i cut 
 
 
 1 
 
 
 3 
 
 VIII 
 
 1 1 
 
 ft 
 
 
 
 
 
 3 
 
 
 I 2 
 
 IX 
 
 < t 
 
 i, I 
 
 
 < < 
 
 
 2 
 
 e 
 
 5 
 
 I-IX 
 
 i i 
 
 
 all cut 
 
 
 
 no fall 
 
 f 
 
 r s 
 
 L 6 
 
 VIII 
 X-XI 
 
 
 
 
 style cut 
 
 a 
 
 
 5 
 
 no fall 
 
 g 
 
 5 
 
 VIII -XIV 
 
 
 
 
 
 slit 
 
 < i 
 
 
 r 3 
 
 VIII 
 
 
 all cut 
 
 < i 
 
 
 
 4 
 
 h 
 
 5 
 
 
 
 t ( 
 
 1 1 
 
 
 
 no fall 
 
 
 L 4 
 
 II-VIII 
 
 
 a 
 
 
 
 
 < < 
 
396 University of California Publications in Botany [VOL. 5 
 
 indicated above, in the case of injury to the ovary in which this organ 
 may be merely punctured, without necessarily removing any material, 
 yet abscission occurs in one to two days after the injury. 
 
 It has, on the other hand, been evident throughout all the abscis- 
 sion experiments that age of flower is the important factor in deter- 
 mining the reaction time, older flowers nearly always responding more 
 slowly to stimulation by injury than younger ones. It will be seen, 
 however, from the tables that there are occasionally individual excep- 
 tions to the general rule. These exceptions might be explained in a 
 number of ways. For example, it is possible in the case of older flow- 
 ers that the ovary, having increased in size, was accidentally cut in 
 the operation of injury, thus adding the extra factor of stimulation 
 of the ovary which in younger flow r ers would not be present. In gen- 
 eral, such exceptions to the general rule indicate to what extent the 
 normal or abnormal physiological conditions of the plant enter into 
 the problem. 
 
 2. ABSCISSION TIME 
 
 The abscission time, or the actual time involved in the process of 
 cell separation, was considered in a preliminary paper (Goodspeed and 
 Kendall, 1916) wherein the minimum time in which abscission was 
 known to have occurred was stated to be from four to eight hours in 
 normal abscission and from one to four hours in "spontaneous" 
 abscission. A few additional data are now at hand in the case of 
 F x H179 and Nicotiana Tabacum ' ' Maryland. ' ' These two forms, 
 as has already been noted, are a little more sensitive than most 
 Xicotiana varieties and normal abscission was found to take place in 
 from three to six hours. 
 
 The time of cell separation in "spontaneous" abscission can be 
 more exactly determined than that in normal abscission because of 
 the regularity with which the plants respond to certain conditions of 
 injury or to the presence of narcotic vapors. Data on this point were 
 obtained in the following manner. Flowering shoots with flowers of 
 different sizes were cut, placed in water and inserted under a bell-jar. 
 Enough illuminating gas was then introduced under the jar to make 
 1.5 per cent approximately. The temperature during the experiment 
 was practically constant at 19 C. After the shoot had been left in 
 this abnormal atmosphere for five hours a few flowers were picked off 
 at fifteen-minute intervals and free-hand sections made of their 
 pedicels until flowers about the size of those which were being sec- 
 
1918] Kendall: Abscission of Flowers and Fruits in Solan-aceae 397 
 
 tioned began to fall. It was found that signs of abscission hardly 
 ever appeared until thirty to forty minutes before actual fall occurred. 
 This indicates that the actual process of cell separation in F! H179 
 takes place in from thirty to forty minutes. Experiments carried on 
 in the same manner with N. Tabacum "Maryland" indicate that 
 abscission here takes place in from forty-five to sixty minutes. 
 
 Both the reaction time of abscission and the actual abscission time 
 are profoundly influenced by temperature and by humidity. Varia- 
 tion in the intensity of the illumination, however, seems to have no 
 direct influence upon abscission. In comparing the effect of changes 
 in temperature and humidity it was found that the results of experi- 
 ments intended to show the time of abscission are far more dependent 
 upon temperature than upon humidity. This is not because changes 
 in humidity have little influence upon abscission but because such 
 changes have to be very great indeed before bringing about any appre- 
 ciable effect. Very slight changes in temperature, on the other hand, 
 often influence abscission to a marked degree. Abscission goes on 
 very actively under high temperatures and converse^ very slowly 
 under low temperatures. It starts in the case of F x H179 about seven 
 hours after insertion in 1.5 per cent illuminating gas at a temperature 
 of 19 C. If the same experiment be repeated in a temperature of 
 approximately 9 C. abscission may not occur for fifteen to twenty- 
 four hours. 
 
 Drought has to be quite severe before retarding abscission. There 
 is 110 doubt, however, that wilted shoots will not drop flowers as 
 quickly as fresh ones and if the wilting proceeds far enough no abscis- 
 sion will occur. This effect is all the more noticeable if the air around 
 the wilted shoot is kept free from moisture. 
 
 EXPERIMENTAL INDUCTION OF ABSCISSION 
 1. INDUCTION BY ILLUMINATING GAS 
 
 The first subject to be considered under this heading is the com- 
 parative effect of illuminating gas in causing abscission in several 
 species of the Solanaceae. The method of determining this consisted 
 largely in placing flowering shoots of the different species in water 
 under bell-jars and introducing enough illuminating gas under the 
 jars to make the percentage of narcotic vapors in the air around the 
 plant 1.5. The temperature during the experiments was compara- 
 
398 University of California Publications in Botany [VOL. 5 
 
 tively high, ranging from 15 to 20 C. The results, which were 
 recorded approximately fifteen hours after subjection to the gas, arc 
 given in the following table : 
 
 TABLE 6 
 
 Species, variety, or Amount of abscission, expressed almost entirely 
 
 hybrid in terms of size of flowers thrown off 
 
 N. Tabacum var. macrophylla all buds up to anthesis. 
 
 N. Tabacum " Maryland" all flowers up to 4 or 5 days past anthesis. 
 
 F, H154 all buds up to opening of corolla. 
 
 F, H36 all buds and flowers. 
 
 F,H179 all buds and flowers. 
 
 N. glauca young buds. 
 
 N. rustica var.? buds up to anthesis. 
 
 N. rustica var.? buds, flowers, and fruits. 
 
 N. Bigelovli var. Wallacei no abscission. 
 
 N. Bigelovii "Porno" no abscission. 
 
 N. quadrivalvis no abscission. 
 
 N. multivalvis no abscission. 
 
 N. Sanderae buds up to anthesis. 
 
 N. suaveolens buds up to anthesis. 
 
 N. plumbaginifolia buds up to opening of the corolla. 
 
 Solanum umbelliferum ^ small buds. 
 
 S. jasminioides buds and flowers. 
 
 S. verbascifolium no abscission. 
 
 S. nigrum small buds. 
 
 lochroma tuberosa no abscission. 
 
 Oestrum fasciculatum buds and flowers. 
 
 Lycopersicum esculentum var. 
 
 pyriforme no abscission. 
 
 L. esculentum var. vulgare small buds and occasional flowers. 
 
 Petunia hybrida no abscission. 
 
 Salpiglossis sinuata -no abscission. 
 
 Datura sanguineum buds and flowers. 
 
 Salpichrora rhomboidea no abscission. 
 
 Lycium australis - no abscission. 
 
 As might be expected, most of these varieties react to laboratory 
 air in the same manner that they do to illuminating gas. In the case 
 of laboratory air a longer time and higher temperature is generally 
 required before the reaction occurs. All the species, with the excep- 
 tion of those which throw off only young buds, detach most of their 
 flowers when left in laboratory air overnight. If a window or two is 
 left open, allowing fresh aid to enter and at the same time lowering 
 the temperature, no abscission occurs. 
 
 It was found that several of the species recorded above, in which 
 no abscission or very little abscission occurred, detached more flowers 
 when a larger percentage of gas was used or when subjected to 1.5 
 per cent gas for a longer time. Thus, both varieties of Lycopersicum 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 399 
 
 esculent um, lochroma tuberosa, Solanum nigrum, and 8. verbasci- 
 folium, upon subjection to 3 per cent illuminating gas for twenty 
 hours, throw off all flowers up to those two or three days past anthesis. 
 No abscission occurred, however, in any concentration of gas, in 
 Mcotiana Bigelovii, N. quadrivalvis, N. multivalvis, Lycium australis, 
 Petunia hybrida, Salpiglossis stinuata, or Salpichrora rhomlioidea. 
 
 A peculiar condition exists in Solanum umbelliferum, which throws 
 off buds in the illuminating gas but never under any conditions, in- 
 cluding temperature or the presence of narcotic vapors, throws off 
 flowers in which the corolla has fully opened. A corresponding con- 
 dition seems to exist in Nicotiana Tabacum var. macrophylla, FJH154, 
 N. Sander ae, N. rustica var. brasUia, and in one other variety of N. 
 rustica, all of which seldom under any conditions detach fully opened 
 flowers, although flowers up to that stage are freely abscissed. Thus 
 there seems to be, in certain species and at about the time of the open- 
 ing of the corolla, a sudden increase in resistance to the external 
 stimulus which is causing abscission. In other species this sudden 
 increase in resistance does not take place, abscission commonly occur- 
 ring at any stage in the development of the flower or fruit and the 
 increase in resistance taking place very gradually. In addition, there 
 seems to be an intergradation of forms between those in which the 
 increase in resistance takes place suddenly and those in which it takes 
 place gradually. 
 
 The next subject to be taken up is a consideration of experiments 
 5, 6, 7, 8, and 9 on the induction of abscission in small isolated pieces 
 of the pedicel. The main purpose of devising these experiments was to 
 throw some light, if possible, on the direct or indirect action of the 
 external factor in causing "spontaneous" abscission. The pedicel of 
 F x H179 was again chosen as material for the following experiments, 
 
 Fig. 9 
 
400 
 
 University of California Publications in Botany [VOL. 5 
 
 largely because of the ease and regularity with which abscission is 
 induced in this hybrid by sudden changes in the external environment. 
 
 Experiment 5. This experiment was devised to discover the effect 
 of reducing the volume of material proximal to the separation layer 
 on the abscission of flowers of Nicotiana as induced by illuminating 
 gas. Two series of flowers were cut as in figure 9. In the last two 
 flowers represented on the right the cut was made less than 0.5 mm. 
 from the separation layer. These flowers were then rolled in damp 
 filter paper and left in 1.5 per cent illuminating gas overnight. After 
 fifteen hours, abscission had occurred in all the flowers except the one 
 represented on the extreme right in the figure. Abscission had 
 occurred in one flower in which the cut had been made less than 
 0.5 mm. from the separation layer. The control to this experiment 
 showed that abscission does not occur for several days in a series of 
 flowers cut as in figure 9 and kept under normal conditions. 
 
 Experiment 6. This experiment was devised to show the effect 
 upon abscission of reducing the volume of material distal as well as 
 proximal to the separation layer. In this case the flowers were cut off 
 at varying distances from the separation layer, making the series 
 shown in figure 10. The last two pieces on- the right in this series 
 were cut less than 0.5 mm. on each side of the separation layer so 
 that the total length of the pieces was not much above 1 mm. In this 
 experiment and in similar ones which follow it was necessary to keep 
 
 Fig. 10 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 401 
 
 the material moist. This was accomplished in various ways, but the 
 best method was found to consist in placing the pieces on a long strip 
 of filter paper one end of which rested in water. In this experiment 
 abscission occurred after ten hours subjection to 1.5 per cent illum- 
 inating gas in all except the two pieces represented in the extreme 
 right of figure 10. Abscission here took place in several pieces rang- 
 ing from 1 mm. to 2 mm. in length. A microscopic examinatiDimf the 
 separation surfaces indicated that the process of abscission corre- 
 sponded entirely with normal abscission as it occurs in plants in the 
 field. Experiments made in a similar manner upon N. Tabacum 
 "Maryland" and Lycopersicum gave similar results. In the control, 
 which consisted in keeping pieces of the pedicel as shown in figure. 10 
 under normal atmospheric conditions, abscission occurred after about 
 twenty hours, evidently as the result of no other stimulus than that 
 caused through cutting oft 3 the flower by severing the pedicel. The 
 reaction in the control, however, is much slower than in the case in 
 which the added effect of the illuminating gas is operative, indicating 
 that the latter factor, although it here serves merely to hasten the 
 abscission process, has an effect of some kind on the tissues at the 
 base of the pedicel. 
 
 Following these two experiments, a number of attempts were made 
 in the same way to induce abscission in longitudinal free-hand sections 
 of the pedicel cut for microscopical examination. It was soon discovered 
 that the abscission process could be induced in the separation zone in 
 thick longitudinal sections of the pedicel by subjecting them to high 
 percentage (5 to 7 per cent) of illuminating gas. Cell separation in 
 cross-sections through the separation zone could not be induced by any 
 means at hand. The following experiments give more detailed results 
 in this connection. 
 
 Experiment 7. In this experiment, median, longitudinal sections 
 of varying thickness were cut through the pedicels so that the plane 
 of the sections corresponded with the plane formed by both the 
 pedicel and the main axis of the inflorescence. These sections were 
 subjected to 7 per cent illuminating gas, care being taken to keep them 
 moist, but not submerged, throughout the entire experiment. The 
 best arrangement was found to be one in which the sections rested in 
 a thin film of water on one side but were exposed to the air on the 
 other. After several hours in the 7 per cent illuminating gas, abscis- 
 sion started in the thicker sections but not in the thinner ones. The 
 extent to which abscission proceeded depended upon the thickness of 
 
402 University of California Publications in Botany [VOL. 5 
 
 the section. Abscission became complete in sections 0.3 mm. or more 
 in thickness, the separation taking place in such a way that a slight 
 bending or pulling motion sufficient to break the trachea divided the 
 section into equal halves. In thinner sections, ranging from 0.3 mm. 
 to 0.17 mm., abscission starts in the normal position but does not pro- 
 ceed to completion, the extent to which the process takes place depend- 
 ing, as has been said, upon the thickness of the section. In sections 
 much below 0.17 mm. no signs of abscission appear. Also, if the 
 thicker sections are shortened in length to any considerable extent by 
 cutting off portions of the tissues from either side of the separation 
 layer, abscission will not occur. 
 
 The process of abscission as it occurs in these sections corresponds 
 exactly to the process in an entire pedicel. Cell separation starts 
 independently in the pith and in the cortex, appearing first in that 
 part of the cortex corresponding to the ventral region of the pedicel 
 where, it will be remembered, abscission starts in the entire flower. 
 When mounting the sections on an object slide for microscopical 
 examination, the isolated cells in the, pith lie in position but can be 
 easily washed out with a small jet of water. In the cortex a break 
 soon appears in the epidermis as the result of manipulation in mount- 
 ing and a cavity is formed at that point as the result of the isolated 
 cells of the cortex floating out in the water. 
 
 Experiment 7 was repeated in the case of Datura with similar 
 results, except that in this case abscission was more active since it 
 involved more cells, a situation which one might be led to expect 
 because of the differences between the two species in the normal abscis- 
 sion of entire flowers. It will be remembered that the separation cells 
 of the cortex in Datura are in no way distinguishable from other 
 cortical cells ; yet even in these sections separation occurs in a definitely 
 predetermined position corresponding entirely with the position in 
 abscission of the entire flower. It was even noticed that abscission 
 started in these sections in the same tissues and in the same manner 
 as in normal floral abscission. 
 
 After the thickness of the sections best adapted to obtaining results 
 had been determined, the following experiment was performed on 
 sections cut from different parts of the pedicel. 
 
 Experiment 8. In this experiment a series of longitudinal sections 
 of the pedicel were cut so that the plane of the sections was at right 
 angles to that of the sections cut in Experiment 7. The first section 
 was tangential, on the ventral side of the pedicel, and contained only 
 the epidermis and a few tiers of cortical cells. Section 2 was also 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 403 
 
 tangential but contained a few tracheae on one surface. Section 3 was 
 more or less radial, containing two strands of vascular tissue on either 
 side. Sections 4 and 5 were similar to sections 1 and 2. On subject- 
 ing these sections to illuminating gas it was noticed that abscission 
 started first in sections 1, 2, and 3, appearing last in sections 4 and 5. 
 This result is exactly parallel with the process as it occurs in normal 
 abscission, where the process starts first in the ventral cortex" and in 
 the pith. 
 
 In passing, mention might be made of the peculiar reaction of the 
 tangential sections 2 and 4, which were made up almost entirely of 
 cortical cells with a few vascular elements on one side. When abscis- 
 sion occurred in these sections, a bending or bowing of the section was 
 alwa.ys noticed. This bending was always such that the tracheal tissue 
 was on the concave side, as if the cells of the cortex had undergone 
 considerable expansion while the cells of the vascular tissue retained 
 their original size. From the work of Richter and others, it may be 
 expected that subjection of portions of plant tissues to illuminating 
 gas would cause an increase in turgor in the cells concerned. Thus, it 
 is probable that the bending of the sections, as described above, is due 
 to the increase in turgor of the cortical cells caused by the narcotic 
 effect of the illuminating gas. The extent of the bending was such 
 that most of the cells in the cortex as well as the separation cells must 
 have been involved in the process. On repeating the above experiment 
 with Datura, a similar bending of the tangential sections was even 
 more pronounced than in Nicotiana. 
 
 Experiment 9. As mentioned above, efforts to induce abscission 
 failed in thin sections. The sections in Experiment 9 were cut so that 
 they were thin in the separation layer but thick on either side. Both 
 surfaces of these sections were thus cut slightly concave so that the 
 sections were thickest at the ends and thinnest in the middle, where 
 the separation zone was located. The sections were then subjected to 
 7 per cent illuminating gas as in Experiment 7. It was not possible 
 to cut very thin free-hand sections of the shape described, but it was 
 demonstrated without a doubt that abscission occurred in sections of 
 this peculiar shape which were thinner in the separation zone than 
 thoso in Experiment 7 where abscission had failed to occur. 
 
 " Certain conclusions which can be drawn from experiments 5, 6, 7, 
 8, and 9 are given below. 
 
 1. Abscission can be induced by allowing the external factor to act 
 directly upon the cells in the vicinity of the separation zone (Expts. 
 6, 7, and 8). 
 
404 University of California fiiblications in Botany [VOL. 5 
 
 2. Abscission induced by the above methods in isolated pieces must 
 be independent of transportation of material from the rest of the plant. 
 
 3. The fact that abscission cannot be induced in thick cross-sections 
 of the -separation zone shows that cell separation cannot be induced 
 by the action of the external factor directly on the separation cells. 
 
 4. It is necessary that a certain proportion of the tissues of the 
 pedicel be in intercellular connection with the cells of the separation 
 zone before cell separation will occur, but this proportion is surpris- 
 ingly small (Expts. 7, 8, and 9). 
 
 5. There is evidently increase in turgor in all the cortical cells of 
 the pedicel during abscission induced by the above method (Expt. 8). 
 
 2. ACTION OP ACIDS ON THE SEPARATION CELLS OP Nicotiana 
 
 Under this heading a description will be given of the effect of 
 mineral acids on small isolated pieces such as were used in experiments 
 6, 7, 8, and 9. It was stated above (page 364) that by the use of two 
 mineral acids together with several stains, no chemical difference could 
 be detected between the cell walls of the separation cells and those of 
 normal cortical cells. The present work represents an attempt to 
 determine, by experimental means and by watching through the micro- 
 scope the action of acids on cell walls, whether the cell membranes 
 of the separation cells are more subject to hydrolysis than those of 
 normal cortical cells. 
 
 Experiment 10. Small pieces of the pedicel were prepared as in 
 figure 10. These pieces were boiled for one or two minutes in 4 per 
 cent hydrochloric acid and then washed in water. Upon examination 
 it was found that the pieces could be separated into halves through 
 the separation zone by a slight pulling or bending motion. Microscopic 
 examination of the separation surfaces showed that the break through 
 the cells of the separation zone had taken place along the plane of the 
 middle lamellae of their walls. This same type of separation was 
 brought about without boiling when 10 per cent nitric or hydrochloric 
 acid was allowed to act on the pedicels for approximately five minutes. 
 When longitudinal sections are used in place of entire pedicles, the 
 same results are obtained but much more rapidly. It was also noticed 
 that separation under these latter conditions takes place more quickly 
 in younger pedicels than in older ones. In the pedicels of fully 
 developed fruits no separation could be induced, but in those of 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 405 
 
 immature fruits separation occurred in the cortex but failed to take 
 place within the vascular cylinder. 
 
 Experiment 10 at first glance would seem to indicate that the cell 
 walls of the separation cells are more subject to hydrolysis than normal 
 cortical cells. Another interpretation is possible, however. Actual 
 separation which takes place through the separation zone may be due 
 to the fact that the cells in this zone are small and have "a tendency 
 to be isodiametric, whereas the remaining cells of the cortex are larger 
 and are elongated parallel to the long axis of the pedicel. Hydrolysis 
 of the cell walls may go on with equal rapidity in all the cortical cells 
 at the base of the pedicel, yet upon bending or pulling separation may 
 take place through the region of isodiametric cells because of the inter- 
 locking of the elongated cells in the rest of the cortex. An attempt 
 was made to gain further evidence on this' point by observing through 
 the microscope the action of acids on the cell walls of the tissues con- 
 cerned. "When the action of the acids is thus observed, the walls are 
 seen to soften and to swell to two or three times their normal thick- 
 ness. This effect is all the more noticeable if the walls initially are 
 comparatively thick. Now, since the cells of the separation zone are 
 small and somewhat collenchymatous, or at least have thicker walls 
 than normal cortical cells, the process of swelling in the cell wall is 
 most conspicuous in that region. Indeed, hardly any swelling can be 
 perceived as a result of the acid treatment in the cell walls of normal 
 parenchyma cells of the cortex. However, when a form such as 
 Lycopersicum is examined in which there is a distinct layer of col- 
 lenchyma beneath the epidermis for the entire length of the pedicel, 
 this collenchyma appears to be affected at the same time and in the 
 same manner as the cells of the separation zone of Nicotiana. Also 
 in Nicotiana there seems to be a certain amount of similarity in 
 reaction to acids between the smaller cells of the cortex just beneath 
 the epidermis and those of the separation zone. The conclusion can 
 thus be drawn that the cell walls of the separation cells are no more 
 readily hydrolyzed than those of normal collenchymatous tissues. Of 
 course, the fact still remains that the collenchyma of the cortex may 
 be more subject to hydrolysis than the cortical parenchyma. Now 
 the small cells of the separation zone not only extend across the base 
 of the pedicel but also spread throughout the general region at the 
 base of that organ; it was therefore noticed that the swelling of cell 
 walls was by no means confined to cells of the separation layer but 
 was more or less prominent throughout the w y hole general region at 
 the base of the pedicel. 
 
406 University of California Publications in Botany [VOL. 5 
 
 The general results of these observations are in a sense negative 
 and seem to indicate that the walls of the separation cells are no more 
 subject to hydrolysis than the walls on either side. This, of course, 
 does not preclude the possibility that a difference exists which is too 
 slight to be detected. It appears, however, that the general region 
 at the base of the pedicel may be more subject to hydrolysis than the 
 more distant portions. 
 
 3. INDUCTION BY MECHANICAL INJURY 
 
 The results of experiments on the induction of abscission by mechan- 
 ical injury are recorded in tables 2, 3, 4, and 5, which have already 
 been considered under the heading, "Time of Abscission" (page 384). 
 Several facts of interest brought out by table 2, which deals with 
 Nicotiana Langsdorffii var. granctiflora, are summarized below. 
 
 1. It appears that removal of or injury to the capsule does not 
 cause abscission in mature fruits (table 2, a, 5, and h; table 3, c and 
 d). The same types of injury generally do cause abscission in im- 
 mature fruits. 
 
 2. It seems that a transverse cut completely through the flower at 
 the distal end of the calyx causes abscission only in buds or flowers 
 near anthesis (table 2, c). It appears, however, that such a cut 
 proximal to the distal end of the calyx causes abscission in flowers 
 several days past anthesis as well as in buds (table 2, a, &). 
 
 3. Removal of the entire calyx causes fall in very young buds only 
 (table 2, d). 
 
 4. It seems that slitting both the corolla and calyx longitudinally 
 on both sides from tip to base does not induce abscission even in young 
 buds (table 2, e). 
 
 5. Entire removal of the style or stamens causes fall only in young 
 buds (table 2, / and g). 
 
 6. It appears that injuries to the pedicel do not cause abscission, 
 provided the flower is not entirely cut away (table 2, i). Just here it 
 is worth mentioning that two of the pedicels cut .transversely as 
 recorded in table 2, i, were cut so deep that the flowers bent over 
 and hung only by a few vascular strands and cortical cells. The 
 wound healed over, however, and the two flowers matured with the rest. 
 
 7. It is evident that injuries which reach the ovary are much more 
 effective in causing abscission than injuries affecting the other parts 
 of the flower (table 2, b and c). 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 407 
 
 8. Fertilization has no influence whatever in preventing abscission 
 when the latter is induced by a transverse cut completely through the 
 flower at the base or middle of the calyx (table 3, c and d). 
 
 9. Certain types of injury, such as entire removal of the calyx and 
 stamens or removal of the entire calyx and half the corolla, evidently 
 cause abscission only by preventing fertilization (table 3, a and &). 
 
 Taking up now the results given in table 4, which "dealt with 
 Fj H179, it will be seen that this hybrid is more sensitive to injury 
 than is N. Langsdorffli. Nevertheless, it is very plain that the general 
 conclusions announced above for this latter species hold for F t H179 
 also. There follows a partial summary of the results in table 4 and 
 a comparison of these results with those obtained in the experiments 
 on N. Langsdorffii. 
 
 1. It seems that removal of the calyx causes fall of much larger 
 buds than in N. Langsdorffli (table 4, d). 
 
 2. F! H179 is evidently much more sensitive in its abscission re- 
 action to a transverse cut through the flower at the middle of the calyx 
 than N. Langsdorffii (table 4, a). 
 
 3. It would seem that slitting the calyx and corolla even to the 
 extent of dividing these organs into four longitudinal strips does not, 
 as a general rule, cause abscission. Such an injury does cause abscis- 
 sion only in extremely small buds (table 4, gr). 
 
 4. It appears that puncturing the calyx, corolla and ovary so 
 that a hole is formed about 2 mm. in diameter in the latter organ 
 causes fall in flowers of all sizes up to two or three days past anthesis 
 (table 4, 7i). Since it is evident that such a hole through the calyx 
 and corolla alone would not cause abscission (table 4, #), abscission 
 in this case must be induced by injury to the ovary. 
 
 5. It is evident that a slit completely through the pedicel for its 
 entire length fails to cause fall in buds or open flowers, but where an 
 effort is made to destroy completely the connection between the flower 
 and stem abscission will occur (table 4, i). 
 
 6. Removal of the style or stamens, as a general rule, causes fall 
 only in young buds, but removal of the former organ is probably more 
 effective in causing flower-fall than removal of the stamens (table 4, e 
 and /). On the other hand, where half the corolla is removed along 
 with the stamens fall occurs in larger buds than where only the latter 
 organs are removed (table 4, 6). 
 
 7. Removal of only half the corolla apparently does not induce 
 abscission (table 4, c). 
 
408 University of California Publications in Botany [VOL. 5 
 
 8. Mature capsules of F H179 are apparently more sensitive to 
 injury than those of N. Langsdorffii (table 4, j). 
 
 The table dealing with the experiments on Lycopersicum indicates 
 that flowers of this genus are remarkably resistant to injury, fall 
 occurring only as the result of stimulation when the ovary is injured 
 (table 5, c and d). Since a large number of tomato flowers are nor- 
 mally abscissed from the different inflorescences on a plant, the sev- 
 eral exceptions to the above statement noted in the table probably 
 demonstrate to what extent the normal physiological condition of the 
 plant affects the matter. It seems to be the opinion of most gardeners 
 who are familiar with the tomato plant that floral abscission in this 
 species is more dependent upon soil conditions than upon injury or 
 sudden changes in climatic conditions. It would seem, however, that 
 injuries to very young fruits normally cause fall, but in this case a 
 stage of development is soon reached at which injury to the berry has 
 no effect in inducing abscission (table 5, /). 
 
 Taking the general results of all the experiments into consideration, 
 it is seen, in the first place, that where injury of a certain type causes 
 fall, a stage of development of the flower is soon reached beyond which 
 the injury no longer causes fall. The increase in resistance to the 
 stimulus of mechanical injury takes place gradually in the species 
 investigated, but some of the species are much more resistant than 
 others. In the second place, injuries to the ovary generally cause 
 flower-fall. Thirdly, whether or not flower-fall occurs as a result of 
 injury to other flower parts depends in some way upon the quantity 
 of material removed. Fourthly, injury to the pedicel does not cause 
 abscission unless it breaks entirely the cellular connection between 
 flower and stem. Lastly, it is improbable that fall induced by injury 
 is due to checking the transpiration stream, since injury to the ovary 
 could have no such effect. Also, a cut across the pedicel so that the 
 flower hangs by only a few tracheae must check transpiration from the 
 flower considerably, yet in this case no abscission occurs. 
 
 It was suggested by Bequerel that injury might cause abscission 
 by checking the transpiration stream which passes up through the 
 pedicel. Considerable doubt has already been cast on this point in the 
 above discussion. In order to throw more light on this question the 
 following experiment was performed in an effort to determine whether 
 checking the transpiration stream . of itself and unaccompanied by 
 mechanical injury would cause abscission. 
 
 Experiment 12. As a means of checking transpiration from the 
 flower a coating of paraffin seemed desirable because it hardens 
 
1918] Kendall: Abscission of Flowers and Fruits in Sol-ana-ceae 409 
 
 quickly, thus permitting several coats to be applied. It was doubtful 
 whether other substances, such as lard, cocoa butter or vaseline, which 
 might have been used, would not have been prevented from completely 
 covering the flower in one coating by the presence of numerous hairs 
 and glandular fluid on the calyx. In this experiment flowers were 
 immersed in melted paraffin to within a millimeter of the separation 
 zone and allowed to stand in water under normal atmospheric condi- 
 tions. As a test for abscission, the shoot was shaken or individual 
 flowers tapped from time to time. It was found that several Nicotiana 
 varieties and hybrids differed in their reaction to this treatment as 
 they did in their reaction to illuminating gas. In A T . Tabacum " Mary- 
 land," for example, paraffining the flowers failed to cause abscission 
 for six days, at the end of which time the flowers began to fall, as did 
 those of the control. Some varieties, however, under such treatment, 
 throw off buds at the end of twenty-four hours, but open flowers of 
 the same varieties are never shed. Whether or not the buds fell in 
 these varieties depended largely on the temperature, at lower tempera- 
 tures no fall occurring. Also, in cases where abscission of buds did 
 occur it was evident that something was actually impeding the pro- 
 cess ; none of the white substance formed by the isolated cells was seen 
 at the base of the pedicel and the buds had to be shaken or tapped 
 quite severely before they fell. 
 
 The results of Experiment 12 and the various observations on the 
 induction of abscission by mechanical injury render it extremely 
 unlikely that checking the transpiration stream is ever a direct cause 
 of abscission. The few cases recorded above in which such a condition 
 seems to cause abscission can be better explained by the action of some 
 other factor than that of interference with transpiration. 
 
 In connection with these experiments upon the effect of checking 
 transpiration the results of Lloyd and Balls on the effect of root 
 pruning, etc., in cotton must be mentioned. It was found that a pre- 
 mature shedding of flowers and young bolls followed root pruning 
 and further that, in general, there is a relation between boll-shedding 
 and the rise and fall of the water-table. Proof positive is not sup- 
 plied that root pruning causes fall of flowers by reducing the water 
 supply of the plant body, and any number of other factors may enter 
 in after such mutilation to bring about, in part at least, such a result. 
 Experiments reported in the present paper seem to leave no doubt 
 that, in Nicotiana at least, temperature is a more important factor in 
 controlling abscission than water supply. 
 
410 I'nirersity of California Publications in Botany [VOL.."; 
 
 4. THE ABILITY OF CERTAIN SPECIES TO THROW OFF PEDICELS 
 
 FROM WHICH ALL THE FLORAL ORGANS HAVE BEEN 
 
 REMOVED, AS RELATED TO THE INDUCTION OF 
 
 ABSCISSION BY MECHANICAL INJURY 
 
 It was soon noticed in the experiments that all plants of a species 
 in which floral abscission occurs throw off the remains of the pedicel 
 when this organ is severed at any point distal to the separation layer. 
 If after such an operation no abscission occurs, it can be safety con- 
 cluded that floral abscission never occurs in that species. Petunia 
 hybrida, Salpiglossis sinuata, Salpichrora rhomboidea, and Lyciuin 
 australis are the only species of the list in table 6 which do not absciss 
 flowerless pedicels in this way. Nicotiana Bigelovii, N. quadrivalvis, 
 and A T . multivalvis occasionally do not throw off pedicels under such 
 conditions. The reaction time in cases where the last three species do 
 absciss severed pedicels is very slow (four to fourteen days). 
 
 Turning now to the relation of these observations to the induction 
 of abscission by mechanical injury, it is first necessary to recall the 
 controls used in Experiments 5 and 6 (cf. pages 399 and 400). A fur- 
 ther consideration of the reaction of these controls will suggest that 
 mechanical injury can induce abscission by the action of the stimulus 
 directly on the cells in the vicinity of the separation zone. The con- 
 trol used in Experiment 5, it will be remembered, showed that abscis- 
 sion does not occur under normal conditions in a series of flowers cut 
 as in figure 9. From the control used in Experiment 6 it is evident 
 that merely cutting off the flower at varying distances from the sep- 
 aration layer, forming pieces as represented in figure 10, causes ab- 
 scission to occur, evidently as the result of no other stimulus than 
 that of severing the pedicel. Now, if the cut be made through the 
 pedicel at a point approximately 1 mm. distal to the separation layer 
 in flowers, as represented on the extreme right of figure 9, abscission 
 will occur in the remaining piece, which is now scarcely 2 mm. in 
 length. It is evident that the stimulus caused by severing the pedicel 
 must act directly on the cells in close proximity to the separation 
 zone. Practically the same results are obtained when the transverse 
 cut is made through the base or middle of the calyx. There is no 
 reason to suppose that the stimulus set up by cutting through the 
 flower near the base or middle of the calyx differs in any fashion from 
 that offered by a cut severing only the pedicel. 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 411 
 
 Several interesting conclusions are brought out by an examination 
 of the above facts. In the first place, the abscission of the remains of 
 severed pedicels is probably independent of the transportation of 
 materials from the rest of the plant to the separation zone. It may 
 result from the action of the stimulus directly on the cells in the 
 vicinity of the separation layer and is, therefore, largely independent 
 of such physiological processes as transpiration which might- conceiv- 
 ably enter in. In the second place, abscission induced by mechanical 
 injury is probably of the same nature as that of severed pedicels and 
 therefore probably results from the action of the stimulus directly on 
 the cells in immediate proximity to the separation layer. 
 
 SUMMARY 
 
 The final summary of results given below is presented under 
 several headings corresponding to those of the main body of the 
 paper. Unless otherwise stated, the results given may be taken as 
 applying to all the species of the Solanaceae in which abscission was 
 found to occur. First is presented a complete list of the species which 
 were investigated, indicating by 1 those in which floral abscission 
 never occurs, by 2 those in which it very seldom occurs, and by 3 
 those which were actually examined microscopically to determine the 
 histological structure of the separation zone and the method of 
 abscission. 
 
 3 N. Tabacum var. macrophylla 
 
 3 N. sylvestris 
 
 3 N. Tabacum " Maryland" 
 
 3 FJH154 (N. sylvestris X N. Tab. 
 
 var. macrophylla) 
 
 3 F,H179 (N. sylvestris X N. Ta- 
 bacum "Cuba") 
 3 FjHSG (N. sylvestris X N. Tab. var. 
 
 angustifolia) 
 N. glauca 
 3 N. rustica (2 varieties not bra- 
 
 silia) 
 
 2, 3 N. Bigelovii (3 varieties) 
 2 N. quadrivalvis (2 varieties) 
 2 N. multivalvis 
 X. Sanderae 
 N. rustica var. brasilia 
 X. suaveoleiis 
 
 3 Solanum umbelliferum 
 
 S. tuberosum 
 
 S. jasminioides 
 3 S. verbascifolium 
 
 S. nigrum 
 
 2, 3 lochroma tuberosa 
 3 Oestrum fasciculatum 
 
 Lycopersicum esculentum var. vul- 
 
 gare 
 
 3 L. esculentum var. pyriforme 
 1, 3 Petunia hybrida 
 1, 3 Salpiglossis sinuata 
 3 Datura sanguineum 
 1 Salpichrora rhomboidea 
 1 Lycium australis 
 
412 University of California Publications in Botany [VOL. 5 
 
 HISTOLOGY AND CYTOLOGY OF THE PEDICEL 
 
 1; The separation layer arises in all the species listed above, except 
 Lycopersicum and Solanum faiberosum, at or near the base of the 
 pedicel. In the latter two species the layer is located near the middle 
 of the pedicel, but even in these cases, if one considers the pedicel to 
 be composed of two iiiternodes, the layer occurs at the base of the 
 most distal internode. 
 
 2. The separation layer is preformed, ready to function at any 
 stage in the development of the flower and represents (cf. Kubart's 
 first type, page 350) a portion of the primary meristem which has 
 retained some of its originally active condition. 
 
 3. In all the species except Datura the separation cells are char- 
 acterized by their small size, isodiametric shape, large amount of 
 protoplasm and somewhat collenchymatous appearance. A study of 
 the early histological development of the pedicel indicates that the 
 small size of the separation cells does not necessarily bear any relation 
 to abscission. This statement is supported by the fact that in Datura 
 there is absolutely no visible difference between the separation cells 
 and any other cells of the pedicel. 
 
 4. Various tests with stains, acids, and alkalis fail to indicate any 
 chemical difference between the cell walls of the separation cells and 
 the walls of neighboring cortical cells which do not separate. How- 
 ever, the middle lamellae of cell walls in the general region at the 
 base of the pedicel seem somewhat more easily hydrotysed by acids 
 than in the more distal portions. 
 
 5. A study of the early histological development of the pedicel in 
 Nicotiana and Lycopersicum shows that the grooves near which the 
 separation zone arises do not necessarily bear any relation to abscis- 
 sion. The grooves are formed because, in the development of the 
 pedicel, certain cells do not increase in size so fast as the neighboring 
 cells on either the proximal or distal side. 
 
 6. The development of mechanical tissue in the pedicel of Nicotiana 
 continues through the separation layer, thus frequently holding the 
 fruit on the plant in spite of the fact that abscission commonly occurs 
 in the cortex. In most of the berry-forming species of the Solanaceae 
 this mechanical tissue does not become continuous through the separa- 
 tion layer and thus offers no impediment to fall when abscission occurs 
 in that region. 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 413 
 
 THE PROCESS OF ABSCISSION 
 
 1. The process of abscission conforms to the usual type, which 
 involves the separation of cells along the plane of the middle lamella 
 of the cell wall separating them. 
 
 2. No cell divisions or elongations were observed to accompany 
 abscission. 
 
 3. All the cells across the pedicel in the region of the separation 
 layer take part in separation except the tracheae and cuticle, which 
 must be broken mechanically. The total number of cells which may 
 be involved is greater in some species than in others. This number 
 may also vary in the same species because of changes in the external 
 conditions. 
 
 4. Cell separation is brought about by the hydrolysis and conse- 
 quent dissolution of the middle lamella (primary cell membrane) or 
 perhaps both the primary and, in part, secondary cell membranes. 
 The agency active in the hydrolysis of the cell membranes is probably 
 an enzyme. 
 
 5. An increase in cell turgor frequently occurs during abscission, 
 but probably serves merely to hasten and facilitate the process. Most 
 of the frequently observed expansion and the turgid appearance of 
 the separation cells during abscission are probably due to the natural 
 release of pressure caused by the dissolution of the middle lamellae. 
 
 6 : Abscission of the style and corolla in Nicotiana and Datura 
 resembles, to a large extent, abscission of the flower. 
 
 TIME OF ABSCISSION 
 
 1. The length of time between anthesis and normal flower-fall due 
 to lack of fertilization differs among the varieties of Nicotiana. This 
 variation was found to range between an average of five to eighteen 
 days in some fifteen species and varieties of Nicotiana. A much 
 smaller range of variation (0.7 to four days, with the largest fre- 
 quency in the three day group) was noted for the time between an- 
 thesis and fall of the corolla after pollination. 
 
 2. The stimulation of the stylar tissues by the growth of the pollen 
 tubes tends to shorten the time between anthesis and fall of the 
 corolla, this effect being independent of fertilization. Such stimula- 
 tion of the stylar tissues has no appreciable effect upon floral ab- 
 scission. 
 
 3. Floral abscission occurs in F x H179 seven hours after subjecting 
 shoots of the plant to 1.5 per cent illuminating gas at a temperature 
 
414 University of California Publications in Botant/ [VOL. 3 
 
 of 19 C. It occurs in Nicotiana Tabacum "Maryland" in eight hours 
 under the same conditions. The actual time involved in the process of 
 cell separation in the above-mentioned cases lies within thirty to forty 
 minutes in the hybrid and within forty-five to sixty minutes in the 
 Tabacum variety. Normal abscission in these forms is much slower 
 
 4. The length of the reaction time in cases of flower-fall due to 
 mechanical injury shows that this length of time depends more on 
 the age of the flower than on the type of injury. 
 
 5. Temperature is the most important conditioning factor in esti- 
 mates of the time of abscission. 
 
 EXPERIMENTAL INDUCTION OF ABSCISSION 
 
 1. Floral abscission is induced, in a large number of the species 
 investigated, by illuminating gas or laboratory air. The increase in 
 resistance to abscission stimulated in this manner takes place suddenly 
 in some species, since abscission will not occur after the opening of 
 the corolla. In other species this condition does not exist. 
 
 2. It is possible to induce the process of abscission with illuminat- 
 ing gas in small isolated pieces of the pedicels or in longitudinal sec- 
 tions of the pedicel cut free-hand from fresh material. 
 
 3. Abscission in Nicotiana and Lycopersicum is induced by certain 
 types of severe injury and not by others. Injury to the ovary seems 
 more effective in causing abscission than injury to other parts of the 
 flower. In the case of these other flower parts, it seems necessary that 
 a certain amount of tissue be actually removed or destroyed before 
 fall occurs. Injury to the pedicel does not cause abscission unless it 
 breaks entirely the connection between floral organs and stem. 
 Flower-fall in Lycopersicum is not readily induced by injury. Floral 
 abscission in this genus is more dependent upon physiological condi- 
 tions brought on by abnormal soil conditions. 
 
 4. Experiments on the induction of abscission in small isolated 
 pieces and in flowers with only a small portion of the stem proximal 
 to the separation layer attached indicate that the stimulus produced 
 by the action of external factors such as illuminating gas and mechan- 
 ical injury can cause abscission by acting directly on the cells in close 
 proximity to the separation zone. The action of external factors is 
 thus largely independent of such physiological processes as transpira- 
 tion which might enter in. This statement is supported by experi- 
 ments which show that abscission is not necessarily induced by 
 checking transpiration from the flower. 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 415 
 
 CONCLUSION 
 
 It is proposed in what follows to take up consideration of such 
 phenomena in connection with abscission as are still but slightly 
 understood. One of the most perplexing of these is undoubtedly the 
 definitely predetermined location of the separation layer _when no 
 morphological and sometimes no physiological (Datura) difference can 
 be detected between the cells that separate and those that do not. 
 There need be no doubt, however, that such a difference does exist 
 and that a sufficient refinement of technique will serve to detect it. 
 
 In considering this matter further it may be recalled that the 
 separation layer in axial abscission is located at or near the base of an 
 internode. There is undoubtedly some connection between this fact 
 and the fact that the cells most active physiologically are often found 
 in this region. The growth of an internode may be brought about by 
 the action of an intercalary meristem located at the base of the organ 
 and a meristem so located in some cases retains its original activity 
 in the mature internode. Now it is well known that the walls of young 
 active cells are more readily subject to hydrolysis than the walls of 
 older cells, because of the fact that the former contain more water. 
 If we assume, then, that the internode is a metabolic gradient with 
 the most active cells at the base, it would be expected that the walls 
 of these cells would be more subject to hydrolysis than any other cells 
 of the internode. If some hydrolysing agency becomes active 
 throughout the pedicel, it might be expected that the walls of the 
 cells at the base of the internode would react first, causing their sep- 
 aration and thus cutting off the flower or internode. By assuming 
 in this way that separation always takes place through the most 
 active cells of the internode it seems possible to explain the predeter- 
 mined location of the separation layer. 
 
 There is undoubtedly some connection between the above problem 
 and the fact that some plants must perfect a separation layer before 
 detachment can take place. In such cases the tissues at the base of 
 the organ are too old for separation. The same stimulus which causes 
 abscission in some species causes a renewal of activity at the basal 
 region of an organ, resulting in cell divisions and new cells. These 
 new cells may, under a continuation of the stimulus, separate one 
 from another. 
 
 Another perplexing problem, which also includes many subsidiary 
 problems, relates to the exact course taken by the stimuli in causing 
 
416 University of California Publications in Bota-ny [VOL. 5 
 
 abscission. Experiments described in the present paper have indi- 
 cated that this course may be direct as well as indirect. Assuming for 
 the present that some of the factors bringing about abscission always 
 act directly while others act indirectly, we might classify the general 
 factors operative in the case of the Solanaceae as follows : 
 
 DIRECT 1. Narcotic vapors. 
 
 2. Injury to floral organs. 
 
 3. Sudden rise in temperature. 
 
 4. Lack of fertilization. 
 INDIRECT 5. Changes in soil conditions. 
 
 6. Factors evident in normal physiological development. 
 
 The direct factors act directly on the cells at the base of the pedicel 
 and consequently the reaction time must be comparatively rapid. The 
 indirect factors act indirectly through the general physiological con- 
 dition, which in turn furnishes the direct stimulus for cell separation. 
 In the latter case the reaction time must, as a general rule, be slow. 
 The nature of factors under 6 are most difficult to understand. An 
 example of the action of these factors would be given in those cases 
 where most of the flowers of an inflorescence are normally abscissed 
 leaving only one or two to continue development, and in those species 
 which absciss male flowers after anthesis. 
 
 A further analysis of the course of the abscission reaction intro- 
 duces another unsettled problem the nature of the agency which is 
 directly responsible for the dissolution of the middle lamella. It has 
 been pointed out before that an enzymatic body of some kind is prob- 
 ably involved. The following discussion brings out certain facts 
 which it is necessary to take into consideration when speculating as 
 to the nature of this supposed enzyme. The activity of the enzymatic 
 body must be subject to both internal and external conditions. The 
 enzymatic material must also be extremely sensitive to slight changes 
 in the normal environment. It must be continually present in the 
 cells of the separation zone and ready at any moment to react to such 
 changes in the environment. A comparison of several species in 
 regard to their abscission reactions to the factors listed above indicates 
 that this supposed enzyme must be more sensitive in some species 
 than in others. Indeed, in certain species in which no abscission 
 occurs the enzyme must be absent from the region of the separation 
 zone or entirely inactive. Finally, it seems necessary to assume that 
 in certain species the action of the enzyme is suddenly inhibited at 
 about the time of the opening of the corolla. 
 
1918J Kendall: Abscission of Flowers and Fruits in Solanaceae 417 
 
 It has been noticed in all the experiments detailed above that 
 older flowers are less subject to "spontaneous"- abscission than 
 younger ones. The transition line as to size or age beyond which 
 no abscission occurs can not in most cases be definitely drawn; that 
 is to say, the development of a resistance to stimuli takes place grad- 
 ually. This is probably explained by the fact that cell walls gradually 
 become less subject to hydrolysis with age. The celluloses ad pec- 
 toses lose water with age and it is well known that these compounds are 
 subject to hydrolysis in proportion to the amount of water they con- 
 tain. In those cases where the increase in resistance to stimuli takes 
 place suddenly it is necessary, as suggested above, to assume some 
 kind of inhibitor of the enzymatic action. 
 
 The effect that pollination has in hastening abscission of the 
 corolla is a subject which is related to the phenomena described by 
 Fitting (1909) for orchids. The phenomena are as yet only slightly 
 understood. The explanation seems to involve some relaying of 
 stimulus from cell to cell. This is also involved in the explanation of 
 floral abscission induced by injury to the ovary. These two cases 
 and others indicate that in some instances, at least, abscission responses 
 are related to tropistic responses as Fitting (1911) has suggested. 
 
 Finally, attention may be called to the fact that the most pressing 
 need in connection with all the problems mentioned above is, in the 
 first place, to establish by some experimental means a definite connec- 
 tion between some enzymatic body and the process of abscission and, 
 in the second place, more definite knowledge as to the role which cell 
 turgor plays in cell separation. Taking all the facts into considera- 
 tion, it is evident that abscission is fundamentally a physiological 
 problem, the crux of which lies, as in all such problems, in the bio- 
 chemistry of the cell. 
 
 The studies reported upon above were carried on under the direc- 
 tion and supervision of Professor T. H. Goodspeed and I am under 
 deep obligation to Professor F. E. Lloyd for many valuable sugges- 
 tions both throughout the course of the experiments and during the 
 preparation of this report of them. 
 
418 University of California Publications in Botany [VOL. 5 
 
 LITEEATUEE CITED 
 
 ATKINS, W. E. 
 
 1916. Some recent researches in plant physiology, p. 64. 
 
 BALLS, W. 
 
 1911. Cotton investigations in Egypt, 1909-1910. Cairo Sci. Jour., vol. 5, 
 p. 221. 
 
 BECQUEREL, W. 
 
 1907. Sur un cas remarquable de autotomie de pedoncle floral de tabac 
 provoque par le traumatism de la corolla. C.-E. Acad. Sci. Paris, 
 vol. 245, p. 936. 
 
 BROWN, H. T., and ESCOMB, F. 
 
 1902. The influence of varying amounts of carbon dioxide in the air on 
 photosynthetic process of leaves and the mode of growth. Proc. 
 Eoy. Soc. London, vol. 70, p. 97. 
 
 CORRENS, C. 
 
 1899. Vermehrung der Laubmoose. Jena, 1899. Quoted from Lloyd (1914a). 
 
 EAST, E. M. 
 
 1915. Phenomenon of self -sterility. Am. Nat., vol. 49, p. 77. 
 
 FITTING, H. 
 
 1909. Die Beinflussung der Ochideenbluten durch die Bestaubuiig und durch 
 andere Umstande. Zeitschr. Bot., vol. 1, p. 1. 
 
 1911. Untersuchung iiber die vorzeitige Entblatterung von Bliiten, Jahrb. 
 wiss. Bot., vol. 49, p. 187. 
 
 GOODSPEED, T. H., and KENDALL, J. N. 
 
 1916. An account of the mode of floral abscission in the F x species hybrids 
 
 of Nwotiana. Univ. Calif. Publ. Bot., vol. 5, no. 10, p. 293. 
 
 GORTNER, E. A., and HARRIS, J. A. 
 
 1914. On axial abscission of Impatiens Sultam as the result of traumatic 
 stimuli. .Am. Jour. Bot., vol. 1, p. 48. 
 
 HANNIG, E. 
 
 1913. Untersuchung iiber das Abstossen von Bliiten u.s.w., Zeitschr. Bot., 
 vol. 5, p. 417. 
 
 HOEHNEL, F. E. 
 
 1878. Ueber den Ablosimgvorgang der Zweige einiger JtHolzegewachse und 
 
 seine antomischen Ursachen. Mitteil. forstl. Versuch. Oester., vol. 1, 
 
 no. 3; vol. 3, no. 2. 
 KUBART, B. 
 
 1906. Die organische Ablosung der Korollen nebst Bemerkung iiber die 
 
 Molsche Trennungschichte. S.-B. Akad. Wien, Math-nat. Kl., vol. 
 
 115.1, p. 1491. 
 LLOYD, F. 
 
 1914a. Abscission in flowers, fruits and leaves. Ottawa Nat., 1914. 
 
 1914&. Injury and abscission in Impatiens Sultani. Quebec Soc. f. protection 
 
 of plants, 1914, p. 72. 
 
 1916a. Abscission in Mirabttis Jalapa. Bot. Gaz., vol. 61, p. 213. 
 1916&. Abscission of flower buds and fruits in Gossypium and its relation to 
 
 environmental changes. Trans. Eoy. Soc. Canada, vol. 10, p. 55. 
 
1918] Kendall: Abscission of Flowers and Fruits in Solanaceae 419 
 
 LEE, E. 
 
 1911. Morphology of leaf -fall. Ann. Bot., vol. 25, p. 51. 
 
 LOEWI, E. 
 
 1907. Blattablossung und verwandte Erscheinungen. Proc. Akad. Wien, 
 
 Math-nat. Kl., vol. 166, p. 983. 
 MOHL, H. 
 
 1860. Ueber den Ablosungsprozess saftiger Pflanzenorgane. Bot. Zeit., vol. 
 
 18, p. 273. 
 BEICHE, C. 
 
 1885. Ueber anatomische Veranderungen welche in den perianthkreisen der 
 Bliiten wahrend der Entwicklung der Frucht vor sich gehen. Jahrb. 
 wiss. Bot., vol. 16, p. 630. 
 
 HTER, O. 
 
 1908. Ueber Turgorsteigerung in der Atmospher von Narkotica. Lotos, 
 
 vol. 56, p. 105. 
 
 EICHTEE, O., and GRAFE, V. 
 
 1911. Ueber den Einfluss der Narkotika auf die chemische Zusammensetzung 
 von Pflanzen. S.-B. Akad. Wien, Math-nat. Kl., vol. 120.1, p. 1187. 
 
 STRASBURGER, E. 
 
 1913. Das botanisehe Praktikum, p. 349. 
 TISON, A. (quoted from LLOYD 1914a). 
 
 1900. Eecherches sur la chute des feuilles chez les dieotyledones. Mem. Soc. 
 Linn. Normandie, vol. 20, p. 125. Quoted from Lloyd (1914a). 
 
 WlESNER, J. 
 
 1871. Untersuchung iiber die herbstliche Entblatterung der Holzgewachse. 
 
 S.-B. Akad. Wien, Math-nat. Kl., vol. 64, p. 456. 
 1905. Ueber Frostlaubfall. Ber. Deutseh. Bot. Ges., vol. 23, p. 49. 
 
PLATE 49 
 
 Fig. 1. Base of pedicel of Nicotiana bud showing groove, separation zone, 
 and process of abscission well under way in dorsal cortex. 
 
 Fig. 2. Portion of cortex in the separation layer of Nicotiana showing the 
 bulging of the epidermis, one of the first signs of abscission. 
 
 [420] 
 
UNIV. CALIF. PUBL. BOT. VOL. 5 
 
 [KENDALL] PLATE 49 
 
 Fig. 1 
 
 Fig. 2 
 
PLATE 50 
 
 Fig. 1. Portion of the base of the pedicel of Nicotiana at a late stage in 
 the process of abscission showing the independent origin of the process in the 
 pith. 
 
 Fig. 2. Portion of the cortex in the separation layer of Nicotiana showing- 
 separating cells next to the vascular system. 
 
 [422] 
 
UNIV. CALIF. PUBL. BOT. VOL. 5 
 
 [KENDALL] PLATE 50 
 
 Fig. 1 
 
 
 
 
 
 * 
 
 Fig. 2 
 
PLATE 51 
 
 Portion of the separation layer of Nicotiana showing cells in the process of 
 separation in the upper part of the section. 
 
 I 424 ] 
 
UNIV. CALIF. PUBL. BOT. VOL. 5 
 
 [KENDALL] PLATE 51 
 
PLATE 52 
 
 Fig. 1. Portion of dorsal cortex near the groove in the pedicel of \icot nnm, 
 showing the abscission process well under way. 
 
 Fig. 2. Group of isolated cells washed off from end of a freshly abscissed 
 pedicel of Nicotiana. 
 
 Fig. 3. Single isolated cell showing the thinness of the remaining cell 
 membrane. 
 
 42(5 ] 
 
UNIV. CALIF. PUBL. BOT. VOL. 5 
 
 [KENDALL] PLATE 
 
 - 
 
 i 
 
 '\ 
 
 Fig. ] 
 
 ' 
 
 
 Fig. 3 
 
PLATE 53 
 
 Fig. 1. Portion of pedicel of Lycopersicum, showing groove and separation 
 zone. 
 
 Fig. 2. Portion of cortex of pedicel of Lycopersicum, showing groove and 
 abscission process fairly well along; cell separation first takes place between 
 only two tiers of cells before spreading to others. 
 
 [ 428 ] 
 
UNIV. CALIF. PUBL. BOT. VOL. 5 
 
 [KENDALL] PLATE 53 
 
 
 
 i 
 
 
 Fig. 2 
 
UNTVESSITY OF CALIFORNIA PUBLICATIONS (Continued) 
 
 6. Contributions to the Knowledge of the California Species of Crusta- 
 ceons Corallines. EL by Maurice B&rstow Nichols. Pp. 549-870; 
 
 plates 10-13. April, 1909 _ . _ Jfi 
 
 1. New Chloropbyceae from California, by Nathaniel Lyon Gardner. Pp. 
 
 371-375; plate 14. April, 1909 J.O 
 
 8. Plantae Mexicanae Pnrpusianae, I, by T. S. Brandegee. Pp. 377-396. 
 
 May, 1909 ;. .15 
 
 Index, pp. 397-400. 
 VOL *. 1010-1912. 
 
 1. Studies in Ornamental Trees and Shrub*, by Harvey Monroe HalL Pp. 
 
 1-74; plates 1-11; 15 text-figures. March, 1910 < _ .78 
 
 2. Gracilariopkila, a New Parasite on Gractlaria confervoidcs, by Harriet 
 
 L. Wilson. Pp. 75-84; plates 12-13. May, 1910 ....;~-,--,..^ _ J.O 
 
 3. Plantae Mexicanae Purpuaianae, II, by T. S. Brandegee. Pp. 85-95, 
 
 May, 1910 - - JLO 
 
 4. Leuvenia, a New Genus of Flagellates, by N. L. Gardner. Pp. 97-106; 
 
 plate 14. May, 1910 . ,10 
 
 6. The Genus Sphaerosoma, by William Albert SetchelL Pp. 107-120; 
 
 plate 15. May, 1910 -. ,15 
 
 6. Variations in Nuclear Extrusion Among the Fueaceae, by Nathaniel 
 
 Lyon Gardner. Pp. 121-136; plates 16-17. August, 1910 . . J.8 
 
 7. The Nature of the Carpostomes in. the Cystocarp of Ahnfeldtia ffigarti- 
 
 noides, by Ada Sara McFadden. Pp. 137-142; plate 18. February, 
 1911 .05 
 
 8. On a Colacodasya from Southern California, by Mabel Effie McFadden, 
 
 Pp. 143-150; plate 19. February, 1911 . .08 
 
 9. Fructification of Macrocystis, by Edna Juanita Hoffman, Pp. 151-158; 
 
 plate 20. February, 1911 . .05 
 
 10. Erythrophyllum dclesserioides J. Ag., by Wilfred Charles Twlss. Pp. 
 
 159-176; plates 21-24. March, 1911 .15 
 
 11. Plantae Mexicanae Purpusianae, HI, by T. S. Brandegee. Pp. 177-194. 
 
 July, 191 1 ._.. -.-.. _ .._ J.5 
 
 12. New and Noteworthy Calif orniaa Plants, L by Harvey Monroe HalL 
 
 Pp. 195-208. March, 1912 _..:.' .15 
 
 IS. Die Hydrophyllaceen der Sierra Nevada, by August Brand, Pp. 209- 
 
 227. March, 1912 -.- . . - ~ - .20 
 
 14. Algae Novae et Minus Cognitae, I, by William Albert Setchell. Pp. 
 
 229-268; plates 25-31. May, 1912 .40 
 
 15. Plautae Mexicanae Purpusianae, IV, by Townshend Stith Brandegee. 
 
 Pp. 269-281. June, 1912 ~ ...._... .19 
 
 16. Comparative Development of the Cystocarps of Antitlmmnion and 
 
 Prionitis, by Lyman Luther Daines. Pp. 283-302; plates 32-34. 
 March, 1913 ...: _ .20 
 
 17. Fungus Galla on Cystoseira and Halidrys. by Lulu May Estee. Pp. 305- 
 
 316; plate 35. March, 1913 _ .10 
 
 18. New Fueaceae, by Nathaniel Lyon Gardner. Pp. 317-374; plates 36- 
 
 53. April, 1913 .75 
 
 19. Plantae Mexicanae Purpusianae, V, by Townshend Stith Brandegee. 
 
 Pp. 375-388. June, 1913 .15 
 
 Index, pp. 389-397. 
 VoL . 1912-. 
 
 1. Studies in Ntcoiiana, I, by William Albert SetcheU. Pp. 1-86. De- 
 
 cember, 1912 - 1.25 
 
 2. Quantitative Studies of Inheritance in Nicotiana Hybrids, I, by Thomaa 
 
 Harper Gobdspeed. Pp. 87-168. December, 1912 _ 1.00 
 
 5. Quantitative Studies of Inheritance in Nicotiana Hybrids, n, by 
 
 Thomas Harper Goodspeed. Pp. 169-188. January, 1913 
 
 4. On the Partial Sterility 6f Nicotiana Hybrids made with N. Sylveslris 
 
 as a Parent, by Thomas Harper Goodspeed. Pp. 189-198. March, 
 
 1913 ao 
 
 5. Notes on the Germination of Tobacco Seed, I, by Thomas Harper Good- 
 
 speed. Pp. 199-222. May, 1913 - .25 
 
 6. Quantitative Studies of Inheritance in Nicotiana Hybrids, HI, by 
 
 Thomas Harper Goodspeed. Pp. 223-231. April, 1915 _ 10 
 
 7. Notes on the Germination of Tobacco Seed, II, by Thomas Harper 
 
 Goodspeed. Pp. 233-248. June, 1915 - .15 
 
 8. Parthenogenesis, Parthenocarpy and Phenospermy in Nicotiana, by 
 
 Thomas Harper Goodspeed. Pp. 249-272, plate 35. July, 1915.. .25 
 
UNIVERSITY OF CALIFORNIA PUBLICATIONS (Continued) 
 
 9. On the Partial Sterility of Nicotiana Hybrids made with N, sylvestris 
 as a Parent. II, by T. H. Goodspeed and A. H. Ayres. Pp. 273-292, 
 plate 36. October, 1916 .20 
 
 10. On the Partial Sterility of Nicotiana Hybrids made with N. sylvestris 
 
 as a Parent. III. An Account of the Mode of Floral Abscission in 
 the F, Species Hybrids, by T. H. Goodspeed and J. N. KendalL 
 Pp. 293-299. November, 1916 05 
 
 11. The Nature of the T^ Species Hybrids between Nicotiana sylvestris 
 
 and Varieties of Nicotiana tabacum, with Special Reference to the 
 Conception of Reaction System Contrasts in Heredity, by T. H. 
 Goodspeed and R. E. Clausen. Pp. 301-346, plates 37-48. Janu- 
 ary, 1917 46 
 
 12. Abscission of Flowers and Fruits in the Solanaceae, with Special 
 
 Reference to Nicotiana, by John N. Kendall. Pp. 347-428, 10 text 
 
 figures, plates 49-53. March, 1918 .85 
 
 VoL 6. 1914- 
 
 1. Parasitic Florideae, X, by William Albert SetcheU. Pp. 1-34, platei 
 
 1-6. April, 1914 .35 
 
 2. Phytomorula regularis, a Symmetrical Protophyte Related to Coelat- 
 
 trum, by Charles Atwood Kofoid. Pp. 35-40, plate 7, April, 1914. .05 
 8. Variation in Oenothera ovata, by Katherine Layne Brandegee. Pp. 41- 
 
 60, plates 8-9. June, 1914 .._ J.O 
 
 4. Plantae Mexicanae Purpusianae, VI, by Townshend Stith Brandegee. 
 
 Pp. 51-77. July, 1914 .25 
 
 6. The Scinaia Assemblage, by WiUiam Albert SetchelL Pp. 79-162, 
 
 platfe* 10-16. October, 1914 .7B 
 
 6. Notes on Pacific Coast Algae. L Fylaiella Postelsiae, n. sp., a New 
 
 Type In the Genus Pylaiella, by Carl Skottsberg. Pp. 153-164, plates 
 17-19. May, 1915 .15 
 
 7. New and Noteworthy Californian Plants, n, by Harvey Monroe Hall. 
 
 Pp. 165-176, plate 20. October, 1915 .15 
 
 8. Plantae Mexicanae Purpusianae VII, by Townshend Stith Brandegee. 
 
 Pp. 177-197. October, 1915 _ _ .25 
 
 9. Floral Relations Among the Galapagos Islands, by A. L. Kroeber. 
 
 Pp. 199-220. March, 1916 .20 
 
 10. The Comparative Histology of Certain Californian Boletaceae, by 
 
 Harry S. Yates. Pp. 221-274, plates 21-25. February, 1916 J50 
 
 11. A Revision of the Tuberales of California, by Helen Margaret Gilkey. 
 
 Pp, 275-356, plates 26-30. March, 1916 - .80 
 
 12. Species Novae vel Minus Cognitae, by T. S. Brandegee. Pp. 357-361. 
 
 May, 1916 _ .05 
 
 13. Plantae Mexicanae Purpusianae, Vni, by Townshend Stith Brandegee. 
 
 Pp. 363-375. March, 1917 15 
 
 14. New Pacific Coast Marine Algae, I, by Nathaniel Lyon Gardner, Pp. 
 
 377-416, plates 31-35. June, 1917 .40 
 
 16. An Account of the Mode of Foliar Abscission in Citrus, by Robert W. 
 
 Hodgson. Pp. 417-428, 3 text figures. February, 1918 10 
 
 VoL 7. 1916- 
 
 1. Notes on the Californian Species of Trillium L. I. A Report of the 
 
 General Results of Field and Garden Studies, 1911-1916, by Thomas 
 Harper Goodspeed and Robert Percy Brandt. Pp. 1-24, plates 1-4, 
 October, 1916 .25 
 
 2. Notes on the Californian Species of Trillium L. IL The Nature and 
 
 Occurrence of Undeveloped Flowers, by Thomas Harper Goodspeed 
 
 and Robert Percy Brandt. Pp. 25-38, plates 5-6. October, 1916 _ J.5 
 
 S. Notes on the Californian Species of Trillium L. III. Seasonal Changes 
 in Trillium Species with Special Reference to the Reproductive Tis- 
 sues, by Robert Percy Brandt. Pp. 39-68, plates 7-10. December, 
 1916 .SO 
 
 4. Notes on the Californian Species of Trillium L. IV. Teratological 
 Variations of Trillium sessile var. giganteum H. & A., by Thomas 
 Harper Goodspeed. Pp. 69-100, plates 11-17. January, 1917 -90 
 
NON-CIRCULATING BO 
 
 374794 
 
 UNIVERSITY OF CAUFORNIA UB1