GIFT OF 
 MICHAEL REESE 
 
LAW AND THEOKY IN CHEMISTRY 
 
LAW AND THEOBY IN 
 CHEMISTRY 
 
 A COMPANION BOOK FOE STUDENTS 
 
 BY 
 
 DOUGLAS CARNEGIE, M.A. 
 
 SOMETIME SCHOLAR AND DEMONSTRATOR IN CHEMISTRY OF GONVILLK 
 AND CAIUS COLLEGE, CAMBRIDGE 
 
 LONDON 
 LONGMANS, GREEN, AND CO. 
 
 AND NEW YORK: 15 EAST 16 th STEEET 
 1894 
 
 All rights reserved 
 

 
 o 
 
 
PEEFACE 
 
 THIS book contains the substance of a summer-school 
 course of eight lectures delivered before an audience 
 of school teachers of elementary chemistry at Colorado 
 College, Colorado Springs, U.S.A. 
 
 It is not therefore adapted to beginners, but it is 
 hoped that it may be found useful by such students 
 as wish to recapitulate and co-ordinate the more im- 
 portant principles of chemistry before proceeding to 
 more detailed and advanced works. 
 
 The qualification companion book occurring on the 
 title-page is used advisedly, for text-book continuity 
 and completeness cannot be claimed for the matter 
 of the following pages. The seven chapters are in 
 reality short and independent essays on the subjects 
 of which they severally treat. 
 
 In the choice of subject-matter I was chiefly in- 
 fluenced by the two following considerations. Firstly, 
 I desired to treat of those subjects which in my 
 opinion are essential to a liberal understanding of 
 the science, but which are inadequately treated in, 
 or altogether crowded out of, the current text-books 
 of elementary chemistry. In the second place, I 
 
vi PKEFACE 
 
 wished to restate and emphasise such points as in 
 my experience present especial difficulties to the 
 student. 
 
 Throughout, the attempt has been made to keep 
 the information up to date, and to indicate, with due 
 appreciation of perspective, the trend of modern 
 research in its relation to the science as a whole. 
 In this connection it is perhaps scarcely necessary 
 to state that the hope was to excite interest rather 
 than to satisfy it. 
 
 I would here express my sincerest thanks to my 
 friend, Mr. Pattison Muir, for the very valuable help 
 he has given me. 
 
 DOUGLAS CAENEGIE. 
 
 BLACKHEATH. 
 
CONTENTS 
 
 CHAP. PAGE 
 
 I. ALCHEMY AND THE BIRTH OP SCIENTIFIC CHEMISTRY 1 
 
 II. THE PHLOGISTIC PERIOD AND THE BEGINNINGS OF 
 
 CHEMICAL THEORY 24 
 
 III. CHEMICAL CLASSIFICATION. MIXTURES, COMPOUNDS, 
 
 ELEMENTS .40 
 
 IV. THE ATOMIC THEORY 69 
 
 V. THE CLASSIFICATION OF COMPOUNDS. ACIDS, BASES, 
 
 SALTS 121 
 
 VI. ISOMERISM AND MOLECULAR ARCHITECTURE . . 148 
 
 VII. CHEMICAL EQUILIBRIUM 186 
 
 vii 
 
LIB/ 
 
 Of TH 
 
 UNIVERSITY 
 
 LAW AND THEOKY IN CHEMISTRY 
 
 CHAPTER I. 
 
 ALCHEMY AND THE BIRTH OF SCIENTIFIC 
 CHEMISTRY. 
 
 IN the ancient world the only scientific pursuits which 
 were considered reputable, and which led to fame or 
 distinction, were politics, philosophy, and mathematics. 
 
 Dyeing, tanning, bronze founding, glass making, 
 soap making are all industries which boast a remote 
 antiquity, but the early practice of these arts was 
 confined to the slaves, and the ignominy of attempting 
 any adaptation of theory to practice never for a moment 
 suggested itself to the noble-born. 
 
 But while the actual practice of what we now under- 
 stand by the term commercial chemistry was in the 
 hands of the uneducated, and possessed no literature,^ 
 the philosophers theorised on the composition and 
 nature of matter, and to trace these theories and their 
 consequences is, as we shall find, to trace the early 
 history of our science. 
 
 The early Greek philosophers of the seventh to the 
 
 A 
 
2 LAW AND THEORY IN CHEMISTRY 
 
 fifth centuries B.C. belonged to one or other of two 
 schools known as the Ionic and the Eleatic schools 
 respectively. The adherents of the Ionic school 
 believed that in all the apparently diverse things by 
 which they were surrounded there existed a common 
 principle or substratum (ap^rf), and they sought to 
 evolve order and simplicity out of the apparent 
 diversity in terms of this common principle. Accord- 
 ing as the attempt was made to investigate and detect 
 this principle by the observation of natural phenomena, 
 or by the unaided exercise of the mind, we find Thales 
 asserting water to be the substratum of all things, or 
 Pythagoras selecting number as the one real thing in 
 nature. The attitude of the Pythagorean branch of 
 the Ionic school is well illustrated by the following 
 quotation from Plato's seventh book of the " Republic " : 
 " We shall pursue astronomy with the help of problems 
 just as we pursue geometry, but if we wish to become 
 thoroughly acquainted with astronomy we will let the 
 heavenly bodies alone." 
 
 The disciples of the Eleatic school founded by 
 Xenophanes did not attempt to explain the material 
 universe, but they espoused a philosophy of an essen- 
 tially negative character. Assuming unity and im- 
 mutability in the universe, without attempting to 
 further specify or explain either the one or the other, 
 they maintain that change and the apparent diversity 
 of things are inexplicable delusions of the senses. 
 Reason was made the only criterion of truth, and as 
 reason apparently led to results at variance with the 
 
THE BIRTH OF SCIENTIFIC CHEMISTRY 3 
 
 information conveyed by the senses, 1 the latter were 
 regarded as fallacious. 
 
 To summarise the differences between the two schools 
 briefly, we may say that while the Ionic school sought to 
 explain the many in the one, the Eleatic school denied 
 the real existence of the many and assumed the one. 
 
 The views of Empedocles (440 B.C.) do not, however, 
 strictly conform to either of the two schools just de- 
 scribed. Unlike the Eleatics, he aimed at a rational 
 explanation of the diversity and change in the universe, 
 and unlike the Ionic school this explanation was given 
 in terms, not of one, but of four common principles, 
 viz., earth, fire, air, and water. These substances, to 
 which Empedocles assigned mythological names, were 
 regarded as the constituents or elements out of which 
 everything else was built up. In Empedocles' opinion, 
 their association in different proportions sufficiently 
 accounted for all the diversities presented by the world 
 to the senses. But to account for the continual 
 changes that are ever taking place in these diversities, 
 Empedocles had to assume the existence of two other 
 rival elements, viz., love and strife, 2 the former causing 
 the attraction and combination of unlike elements, the 
 latter causing their disruption. If we substitute the 
 seventy odd elements of to-day's chemistry for the four 
 Empedoclean elements first mentioned, and then the 
 
 1 Zeno's well-known paradox of Achilles and the tortoise may here 
 be instanced. 
 
 2 We should be inclined to regard the "love" and "strife" of 
 Empedocles rather in the light of powers or influences than as elements. 
 But Burnet (Early Greek Philosophy] states that Empedocles himself 
 regarded them as corporeal elements co-equal with the other four. 
 
4 LAW AND THEORY IN CHEMISTRY 
 
 term affinity for the remaining " love " and " strife/' we 
 cannot but be struck with the close analogy which the 
 views of this early Greek philosopher bear to those held 
 in the nineteenth century concerning the composition 
 of, and the changes undergone by, the material universe. 
 
 The views of Empedocles were modified, and in their 
 modified form popularised by Aristotle (385-322 B.C.). 
 The great Stagy rite regarded all substances as portions 
 of one and the same fundamental matter, modified by 
 the greater or less quantities of the four " elementary 
 principles " earth, fire, air, and water which it had 
 impressed on it. This conception of elementary prin- 
 ciples has nothing in common either with our present 
 conceptions of the elements or with the original views 
 of Empedocles. It is one of those abstract ideas ori- 
 ginally borrowed in all probability from India, where 
 Buddha taught that in addition to the four principles 
 already mentioned, other two ether and consciousness 
 existed, and it is very difficult for us to replace our 
 clear-cut notions of definite concrete elements by these 
 unreal products of the Oriental imagination. 
 
 It must not be supposed that Aristotle intended to 
 convey the idea that water, air, fire, and earth, as we 
 now understand these terms in a purely materialistic 
 sense, constituted the universe by their manifold com- 
 binations. The elementary principle of earth implied 
 to the Aristotelian school merely the possession by the 
 bodies supposed to contain it, of the properties of 
 drvness and coldness. Similarly the principle of water 
 implied coldness and moistness, the principle of fire, 
 
THE BIETH OF SCIENTIFIC CHEMISTRY 5 
 
 heat and dryness, and the principle of steam or air, 
 heat and moistness. 
 
 Accordingly it was held that one specific kind of 
 matter could be totally changed into another by merely 
 altering the proportion in which the principles had 
 been impressed on the fundamental matter of the first 
 substance by altering the ratio of its heat, coldness, 
 dryness, and moistness. Thus by imparting more heat 
 to water (fundamental matter especially richly im- 
 pressed with the principle of water) it becomes steam 
 or air. By taking away its moisture and cooling it, 
 it becomes, as was then taught, rock crystal or petri- 
 fied water, fundamental matter especially rich in the 
 principle of earth. , 
 
 The important point to note is that the Aristotelian 
 doctrine favoured the idea of a possible transmutation 
 of any one substance into any other. 
 
 Experiments having as their aim the change of one 
 substance into another seem first to have been ex- 
 tensively practised in Egypt, which was once called 
 Klimi (Greek, %?7/ua) on account of the darkness of 
 its soil ; Khmi literally meaning black soil. And it 
 came about that the art which had as its aim and 
 object the transmutation of material things received 
 the name of the country where it first flourished the 
 art was called fflpla, 1 just as a certain form of wool- 
 spinning is called worsted-making from the village of 
 Worsted in Norfolkshire, where the industry had birth. 
 
 1 The use of this term to designate the art first appears, so far as 
 we know, in the fourth century. 
 
6 LAW AND THEORY IN CHEMISTRY 
 
 In the seventh century the great wave of Islamism 
 rolled westward from Arabia towards Europe, and 
 the missionaries of Mahomet overran and conquered 
 Egypt. 
 
 During their sojourn in Egypt, the Arabians found 
 the dwellers in the land executing what were then 
 regarded as actual transformations of matter. 1 For in- 
 stance, when molten sulphur was poured into mercury, 
 the metallic properties of the latter disappeared and 
 the mercury and sulphur changed into a solid substance 
 black as the raven's wing. When this black substance 
 was heated gently for some time, it changed into the 
 beautiful red substance now used as a paint under the 
 name vermilion. 
 
 This does not strike us now as a very wonderful 
 change, but owing to the way in which the early 
 naturalists were wont to symbolise things, it was to 
 them a change of the profoundest type. For black 
 was with them emblematic of evil, while red denoted 
 virtue ; and the change from a black substance into 
 a brilliant red one was in their eyes as the miracle 
 of an evil tree bringing forth good fruit. 
 
 But the change which appealed most to the Arabian 
 conquerors was the, to their minds, less miraculous 
 change or transmutation of the base or common metals 
 into the noble metals, gold and silver. Though the 
 
 1 It is probable that prior to their invasion of Egypt, the Arabians 
 had already learnt something of Egyptian arts through the Nestorians. 
 In the fifth century this Christian sect was banished from its home in 
 Constantinople to the desert of Thebais in Egypt, whence its followers 
 gradually migrated eastwards. 
 
THE BIRTH OF SCIENTIFIC CHEMISTRY 7 
 
 modus operandi of this transmutation was kept very 
 secret, 1 yet the" possibility of the transmutation was 
 regarded as a well-ascertained fact during the first 
 centuries of the Christian era. The Arabians carried 
 away with them in their western trend this idea of 
 the ennobling of the base metals, and introduced it 
 into Europe under the title Alchema ; the prefix Al 
 being the Arabic for the definite article, and chema 
 being the Arabic rendering of yyiiia, but signifying 
 dark in the sense of obscure or secret rather than 
 black. In all probability the prefix Al was added 
 for the purpose of conferring a dignity and distinction 
 on the art as being the highest and most reputable 
 of all arts, just as the Spanish discoverers of the New 
 World called the American crocodile el lagarto the 
 lizard because it was the largest and most formidable 
 lizard-like animal they had ever seen. El lagarto 
 and alcherna persist in the language of to-day as 
 the words alligator and alchemy. The word alchemy 
 gave rise by apherisis to the word chemist, and this 
 in turn to the word chemistry. 2 (Of. poet and 
 poetry.) 
 
 1 The earliest chemistry of all is to be found in the ayia r^vv) or 
 holy art, whose secrets were kept sacred under "pain of the peach 
 tree" (i.e., under the penalty of being poisoned by hydrocyanic acid), 
 whose study was confined to the priesthood and the sons of kings, and 
 whose laboratories were the temples. 
 
 2 There are some who, rejecting the derivation of the word Alchemy 
 given in the text, would derive X 7 ?/-"' 01 from %iy/.eia (a mingling, or 
 infusion) from "xyfibs (juice), which in turn is related to x^ iv (t pour). 
 Hence the two current methods of spelling the word chemistry = 
 chymistry. 
 
 Picton in his book The Story of Chemistry rejects both these deriva- 
 
8 LAW AND THEORY IN CHEMISTRY 
 
 The Arabian alchemists, as we learn from the writings 
 of Geber works dating from the eighth century and 
 constituting the oldest chemical literature extant 1 
 had a theory of their own respecting the nature of 
 the metals. Aristotle's theory was wide and applied 
 to all kinds of matter, whereas the alchemistic theory 
 was specialised and had reference more particularly to 
 the metals. The latter theory, moreover, did not re- 
 place the former, but only supplemented it. Aristotle's 
 elementary principles were still recognised, but regarded 
 as the more remote factors in determining the properties 
 of matter, while Geber's elements represented the more 
 proximate constituents of matter in general and of the 
 metals in particular. 
 
 Geber's idea, which met with a general adoption by 
 the alchemists, was that the metals consist of sulphur 
 and mercury in varying proportions. His sulphur and 
 mercury did not, however, mean what these words 
 now connote to us definite chemical individuals with 
 invariable properties. The names had for him an 
 
 tions, and asserts that Alexander of Aphrodisia (a great exponent of 
 the Aristotelian doctrines in the second century A.D.) invented the 
 word chyniike for the operations of the laboratory ; but surely it is 
 unusual to invent words out of one's inner consciousness without any 
 attempt at appropriateness, historical or otherwise, in the resulting 
 invention. Though Alexander may have thrown the word chymike 
 current on the world, it does not follow that in the selection of the 
 word he was uninfluenced by anything save inventive power. 
 
 1 Some philologists would derive the word Gibberish [said to have 
 been originally Geberish] from the alchemist Geber, notwithstanding 
 the fact that Geber's writings, compared with the inflated and 
 unintelligible jargon of other alchemical productions, are models of 
 conciseness and clearness. 
 
THE BIRTH OF SCIENTIFIC CHEMISTRY 9 
 
 abstract meaning ; his sulphur and mercury varied in 
 kind and in properties. Thus gold and silver were 
 very rich in " pure or perfect kinds " of mercury, but 
 the one contained a red sulphur, and the other a 
 white. In short, Gefcer's elements were rathei of 
 the nature of elementary principles than of the 
 nature of elements as we now understand the latter 
 term. 
 
 Since the change in purity as well as the change 
 in relative proportion of these two metallic constituents 
 was assumed to be under the control of the experi- 
 menter, it is obvious that the attempt to transmute 
 the base metals into gold was quite legitimate from a 
 theoretical standpoint. 
 
 Further, it was generally believed that the evolution 
 of the base metals into the noble ones, gold and silver, 
 was a change which was proceeding spontaneously and 
 constantly in nature, and that therefore the alchemists' 
 task virtually consisted in an artificial acceleration of 
 a perfectly natural process. 
 
 Alchemy scarcely merits the dignity of being classed 
 as a science. For some of its followers it was a 
 transcendental philosophy which amounted almost to 
 a religion ; but for the majority it was nothing more 
 nor less than an empiric trade. Its chief service to 
 science lies in the fact that it stimulated experiment 
 blind and thoroughly haphazard experiment perhaps, 
 but still experiment. Its most fatiguing and laborious 
 study was prosecuted by the majority, not so much 
 out of a desire for scientific triumph or for truth's 
 
io LAW AND THEOEY IN CHEMISTRY 
 
 sake, as out of a thirst for that power which the so- 
 called philosopher's stone with its Midas touch was to 
 confer. The alchemists never dreamed of trying to 
 convert gold into lead. 
 
 The philosopher's stone was the name given to a 
 mythical theurgic powder 1 which was to have the 
 power of fermenting millions of times its own weight 
 of fused base metal into gold. The strangest and 
 most diverse ingredients were mixed together in the 
 attempts to prepare this much sought for powder; 
 we read of snails' slime, serpents' teeth, gall stones 
 taken from cats, blood, hair, white of egg, &c., in 
 addition to the never failing ingredient, mercury. 
 The philosopher's stone was generally identified with 
 what in more general terms was called the " One 
 Thing" the perfect form of matter which was to 
 combine in itself all the properties of all other kinds 
 of matter in their highest perfection. Hence we can 
 in a measure account for the many and varied in- 
 gredients which were believed to be necessary to its 
 production. 
 
 Many of the receipts for making the philosopher's 
 stone that have come down to us are perfectly un- 
 intelligible, so allegorical is the language in which 
 they are couched. Sometimes the directions seem 
 fairly explicit, till one comes to the concluding item 
 "add carefully a sufficiency of you know ivliat" 
 
 1 It may here be stated that some philologists find in the generally 
 entertained belief that the " stone " would turn out to be a black 
 powder, a derivation of the word alchemy. See article " Chemie " in 
 Ladenburg's Ilandworterbuch der Chemie. 
 
THE BIRTH OF SCIENTIFIC CHEMISTRY 11 
 
 Several reasons may be assigned for this secrecy and 
 mystery, and sometimes one, sometimes another of 
 these reasons may have been operative. In the first 
 place, the alchemists were undoubtedly desirous of 
 keeping their profession strictly limited as to numbers, 
 a result which would be effected by the adoption of 
 what practically amounted to a private code. Again, 
 there were alchemists whose attitudes towards their 
 profession are sufficiently expressed in the following 
 quotation : 
 
 " If thou shouldst reveal that in a few words which 
 God hath been forming a long time, thou shouldest 
 be condemned on the great day of judgment as a 
 traytor to the majestie of God." 
 
 Finally, there may have been some alchemists who 
 out of vanity or for purposes of self -aggrandisement 
 asserted, with intent to deceive, their knowledge of 
 the transmuting } substance. Obviously it would be 
 expedient for such men to express themselves so 
 obscurely as to render it impossible for others to 
 bring their statements to the test of experiment. 
 
 With the Benedictine monk, Basil Valentine, begins 
 the period of iatro or medical chemistry, a period 
 lasting from the fifteenth to the seventeenth century, 
 during which transmutation was in abeyance. Here- 
 tofore the apothecaries had prescribed purely vegetable 
 preparations only, but now we find mineral specifics 
 contesting the field with them and partially re- 
 placing them. The true goal of alchemy had come 
 to be looked on as the attainment of health, not 
 
12 LAW AND THEORY IN CHEMISTRY 
 
 wealth " a back tough as Hercules " rather than 
 the riches of Solomon and the philosopher's stone 
 was endued with all the attributes of an elixir of 
 life in addition to its now secondary transmuting 
 powers. 1 It was even customary at this period to 
 ascribe the great ages of the patriarchs to their 
 possession of the health-giving stone. 2 
 
 It is probable that the idea of the prophylactic 
 and life - sustaining function of the stone originated 
 in the great propensity of the alchemists for sym- 
 bolism and metaphor. They even went the length 
 of constructing their apparatus symbolically retorts, 
 for instance, being frequently fashioned so as to 
 resemble men, the bulb of the retort being the 
 stomach, the crown of the retort being the head, 
 and the beak being the nose. 3 
 
 1 Van Helmont (1577-1644), the originator of the generic term gas 
 and the discoverer of the existence of several distinct kinds of gases, 
 believed that the universal solvent or "alkahest," not the philosopher's 
 stone, would prove to be the true elixir vitse. 
 
 2 See Jonson's Alchemist, act ii. sc. i., where reference is also made 
 to the belief that such mythological stories as those of the Golden 
 Fleece, the Hesperian Gardens, the Boon of Midas, &c., were "all 
 abstract riddles of the stone." It may be asked how it came about 
 that the patriarchs died, seeing that they were possessed of the elixir of 
 life. The alchemist would have made answer in this wise. Every 
 person has a predestined maximum lease of life a limit which he 
 cannot possibly pass, but a limit to which by reason of disease he 
 seldom attains. Now it was not claimed for the magic stone that it 
 was more mighty than destiny. All it could do would be, by prevent- 
 ing sickness and consequent premature death at three-score years and 
 ten or thereabouts, to enable men to enjoy the maximum lease of life 
 stretching well up into the hundreds. 
 
 3 See Bolton, Trans. Neiv York Acad. Sc., December 1882, March 
 1883. 
 
THE BIRTH OF SCIENTIFIC CHEMISTRY 13 
 
 In further conformity with this propensity, the alche- 
 mists also referred to the base metals as the diseased 
 metals, the lepers, &c., and figuratively spoke of their 
 transmutation into gold as a healing of their diseases. 
 Thus in all probability arose the idea of the alchemist's 
 elixir of life. 
 
 In the blind search for this elixir many valuable 
 drugs were discovered and fairly fully investigated. 
 Valentine busied himself chiefly with the therapeutics 
 of antimony and its compounds, and he very thinly 
 veiled his impatience with and contempt for " the 
 deplorable, putrid, and stinking bag of worms," that 
 failed to see the wonderful virtues of antimony as he 
 himself saw them. 
 
 In addition to his book, The Triumphal Chariot of 
 Antimony, Valentine left behind him a complication 
 in the views which had been current up to his day 
 respecting the nature of the metals. This complication, 
 which consisted in the assertion of the presence in the 
 metals of a third elementary principle, viz., salt, 1 did 
 not inaugurate any material advance in views respecting 
 the composition of matter. This new idea, however, 
 was soon generally accepted, and in contemporary 
 literature we find all kinds of fanciful analogies in- 
 stituted between such diversities as the Trinity, body, 
 soul, and spirit, and the trio salt 0, sulphur ^, and 
 mercury $ 
 
 1 The principle of salt was represented in that portion of the metal 
 which resisted the action of heat, and remained behind as a solid 
 residuum. 
 
14 LAW AND THEORY IN CHEMISTRY 
 
 Valentine had shown a laudable boldness and inde- 
 pendence in ushering in, against strenuous opposition, 
 the reign of iatro-chemistry, but his immediate successor, 
 Philippus Aureolus Theophrastus Bombastus Paracelsus 
 von Hohenheim that strange character, half scientist, 
 half charlatan carried on the crusade with a vehement 
 intrepidity amounting almost to truculence. After 
 forcibly expressing his contempt for the old school of 
 " medicasters " by a public holocaust of the works of 
 the celebrated physicians Galen (second century) and 
 Avicenna (tenth century), he continued to work in the 
 field first opened up by Valentine, and enriched medicine 
 by the knowledge he contributed thereto of many 
 valuable mineral specifics l (corrosive sublimate, tinc- 
 ture of perchloride of iron, &c.), which still figure in 
 the pharmacopoeias of to-day. He also substituted the 
 more or less pure active principles of plants for the 
 crude electuaries of the apothecary, and it is interesting 
 to note that it was probably through his introduction 
 from the East, and liberal use, not of a mineral, but of 
 a vegetable drug, viz., laudandum, 2 that Paracelsus 
 effected those wonderful cures which gained for him 
 his great and widespread reputation. 3 
 
 1 "The medicines are ranged in boxes according to their natures, 
 whether chymical or galenical preparations." Quoted in Johnson's 
 Dictionary. 
 
 2 Laudandum (of which the modern word for opium, viz., laudanum, 
 is a contraction) is the Latin gerundive signifying "meet to be 
 praised." 
 
 3 In his poem " Paracelsus," Browning, while making use of incidents 
 in the life of the chemist as a mise en scene, sketches the career of an 
 ideal Paracelsus. 
 
THE BIRTH OF SCIENTIFIC CHEMISTRY 15 
 
 Early in his career, Paracelsus devoted himself to an 
 experimental study of transmutation, but it was not 
 long before he concluded that it was a chimera. Some 
 of the best minds among his contemporaries and im- 
 mediate successors were doubtless influenced by the 
 conclusions of so great a personality, and renouncing 
 alchemy, devoted themselves to medical chemistry. At 
 any rate the decline of alchemy, properly so-called, 
 dates from the time of Paracelsus. Most of the so- 
 called alchemists of the latter portion of the middle 
 ages were in reality frauds of the lowest order, whose 
 stock-in-trade consisted in a little leger de main, and an 
 unlimited amount of the inflated and highly allegorical 
 language in which the alchemists were wont to give 
 their recipes for the elusive stone, and to impress the 
 lay mind. 
 
 In the memoirs of the Academy of Sciences for 1772, 
 Geoffrey makes an exposd of the ways and means of 
 these pseudo-alchemists. Hollow stirring rods con- 
 taining gold, and temporarily stopped with wax, 
 crucibles with false bottoms previously charged with 
 the precious metals, and composite rods one half gold 
 soldered to the other half iron, and the whole coloured 
 uniformly were among the properties of the craft. 
 Thus equipped, the charlatans practised on the cre- 
 dulity of the public, easily leading it to believe them 
 possessed of clairvoyance, and a supernatural power 
 enabling them to control natural agencies. Telling 
 fortunes by the stars, dispensing "familiars to rifle 
 with at horses and win cups," &c., they pursued a 
 
1 6 LAW AND THEORY IN CHEMISTRY 
 
 royal road to wealth. Instances of such frauds occur 
 throughout mediaeval literature and history. One of 
 Ben Jonson's best constructed plays The Alchemist 
 is based on the despicable chicanery of a " cunning 
 man," well named Subtle, and his confederate, Face. 
 In the Canterbury Tales, again, Chaucer, apparently 
 moved by some sudden resentment, departs from the 
 order of the poem set forth in the prologue to introduce 
 " The Chanones Yemannes Tale." A yeoman or servant 
 falls in with the body of pilgrims on their way to 
 Canterbury, and, after detailing to them the seductive 
 horrors of alchemy, relates the story of some trickery 
 that his late master, the Chanon, had practised on a 
 priest. 
 
 In actual history we find these pseudo-alchemists 
 figuring in the retinues of bankrupt kings and spend- 
 thrift potentates along with chaplains, jesters, and 
 so forth. At the instigation of Edward II., we read 
 that John Cremer, Abbot of Westminster, invited 
 the famous Spanish chrysop&iis, 1 Eaymond Lully, to 
 come over to England for the purpose of replenishing 
 the royal exchequer; and Henry VI. for long enter- 
 tained a firm and costly belief in alchemy, in spite of 
 the fact that Henry IV. had passed a law said to be 
 the shortest in the statute-book making the practice 
 of alchemy a felony. Indeed, most of the European 
 courts so fostered and encouraged alchemy during the 
 
 1 Chrysopceia (literally, gold making) was a synonym of the term 
 alchemy in use during a long period. See Ben Jonson's Alchemist, 
 act ii. sc. i. 
 
THE BIRTH OF SCIENTIFIC CHEMISTRY 17 
 
 fourteenth and fifteenth centuries that the markets of 
 the day were glutted with worthless and counterfeit coin. 
 Several coins and medals, which are claimed to be 
 products of the alchemical art, are still preserved in the 
 Eoyal Cabinet of Coins in Munich, and in the Imperial 
 Cabinet in Vienna. 1 As many of these relics un- 
 doubtedly consist of pure gold and silver, one is forced 
 to conclude either that the bullion from which they 
 were made was produced fraudulently by such tricks as 
 Geoffrey has explained (p. 15), or that the noble metals 
 resulted from the application of cupellation processes to 
 base metals which were believed to be pure, but which 
 in reality carried noble metal. In agreement with the 
 first conclusion, we read that in 1709 a certain retainer 
 alchemist, Domenico Manuel by name, being detected 
 in his knavery, experienced the irony of execution on a 
 gilded gallows at Kiistrin. In support of the alternative 
 conclusion, it may be stated that there is in the British 
 Museum a facsimile of a silver medal made by Becher 
 (see post, p. 24), which was in all likelihood obtained 
 by the cupellation of argentiferous lead supposed by 
 Becher to be pure lead. 2 The medal bears on its 
 reverse the following inscription : 
 
 "Anno 1675 mense Julio Ego J. J. Becher Docter 
 hanc unciarn argenti purissimi ex plumbo arte alchymica 
 transmutavi." 
 
 1 See H. C. Bol ton's Contributions of Alchemy to Numismatics ; 
 also Reyher, De numis quibusdem ex chymico metallo factis. 
 
 2 The original of this interesting medal is in the Vienna Cabinet of 
 Coins. 
 
 B 
 
i8 LAW AND THEOKY IN CHEMISTRY 
 
 On the obverse is the usual representation of Saturn 
 with his wooden leg, scythe, &c. For the alchemists 
 believed the metals to be under the influences of the 
 planets ; and sluggish lead had assigned to it as its 
 patron star the planet Saturn the slowest moving 
 planet with which the alchemists were acquainted. 1 
 
 Becher, regarding the production of his medal as a 
 bond fide case of transmutation, seems to have valued 
 it simply as a chemical curiosity ; for we are told that 
 he expressed himself more interested in the solution 
 of nature's riddles than in the heaping up of wealth. 2 
 
 Even so late as the year 1843 we find one of the 
 pseudo-alchemistic cult, one Francois Cambriel, of 19 
 Judas Street, advertising in a leading paper his readi- 
 ness to teach in a course of nineteen lessons all the 
 secrets of the Hermetic Art. 3 
 
 But that the bond fide alchemists themselves tho- 
 roughly believed in the possibility of transmutation, 
 there can be no doubt. 4 Such facts as the following 
 
 1 The personification of Saturn as a lame man with hour-glass and 
 scythe, &c., is due to the fanciful identification of the Roman Saturnus 
 with the Grecian Cronus during the Hellenising period. 
 
 2 Becher's offer to the Government of Holland to provide it with 
 six millions of golden thalers per annum if it would provide a certain 
 amount of silver and unlimited sea-sand, does not seem to have been 
 taken up in spite of the fact that an experiment made in 1679 is said 
 to have turned out six times more productive than Becher had 
 anticipated ! 
 
 3 So called after the mythical Hermes Trismegistos, the reputed 
 founder of all arts and sciences, and the Grecian representative of the 
 old Egyptian godhead Thoth the deified intellect. 
 
 4 Even such reputed early-day chemists as Davy, Dumas, and 
 Bergman, could not bring themselves to utterly reject the idea of the 
 possibility of transmutation. Peter Woulfe, a fellow of the Royal 
 
THE BIETH OF SCIENTIFIC CHEMISTRY 19 
 
 could not but foster the brightest hopes of men who, 
 satisfied with surface views of things, never questioned 
 what was under the veneer. Iron plunged into a green 
 solution obtained by dissolving certain ores (containing 
 copper) in nitric acid, was itself apparently changed 
 into the more valuable metal copper ; and copper melted 
 with tutty (an impure oxide of zinc) acquired the 
 bright rich yellow colour of gold. And in the 
 metallurgy of the alchemists much importance was 
 ascribed to purely colour changes ; much was supposed 
 to have been achieved in making base metals take the 
 colour of the noble ones. Indeed, many of the early- 
 day alchemists seem to have believed that to give a 
 base metal the colour and sheen of gold was in effect 
 to change the base metal into gold. In accordance 
 with these beliefs, the transmuting substance is often 
 called a tincture (from tingo, I dye). Thus, on a medal 
 of the year 1647 made by one Hofmann, are inscribed 
 the letters T G V L, probably an abbreviation of the 
 sentence tinctures guttce V libram, meaning that five 
 drops of the tincture used had effected the transmuta- 
 tion of a pound of the base metal from which the 
 medal had been prepared. 
 
 Again, in the process now known as cupellation, 
 lead heated in a vessel made of bone ash, slowly dis- 
 appeared and left behind a button of silver. Pyrites 
 treated in a similar way often left a legacy of gold. 
 Of course the noble metals were ab initio present in 
 
 Society of this century, whose name lives in connection with certain 
 pieces of chemical apparatus, was a firm believer in transmutation. 
 
20 LAW AND THEOKY IN CHEMISTRY 
 
 the lead and the pyrites, but the experiments were for 
 the alchemists, who had no notion of what we now 
 understand by chemical homogeneity, cases of indubi- 
 table transmutation. 
 
 When one recalls the modern work on the marvellous 
 effects of "traces of impurities" in altering the pro- 
 perties of large masses of matter, one can scarcely be 
 surprised at the attitude of the alchemists. Instance 
 in this connection Carey Lea's so-called allotropic forms 
 of silver. Carey Lea has recently prepared, among 
 others, a blue pulverulent form of silver soluble in 
 water, and an insoluble variety of the colour of 
 burnished gold. These interesting bodies are not, it 
 is true, pure silver; yet they contain some 98 per 
 cent, of the precious metal, and would have been 
 startling discoveries in the days when transmutation 
 was regarded as a possibility. Eeference should also 
 be made here to the marvellous effects of great in- 
 dustrial significance produced in the properties of iron 
 by the addition of mere traces of such substances as 
 aluminium, tungsten, &c. 
 
 The transition from the vague and uncurbed fantasies 
 of alchemy into the true science of chemistry, was 
 marked by the appearance of a work written in the 
 form of a discussion conducted by a symposium of 
 scientific men, and entitled " The sceptical chymist, or 
 chemico-physical doubts and paradoxes touching the 
 experiments whereby vulgar spagyrists 1 are wont to 
 
 1 Chemistry was often referred to as the spagyric art, from <rirdw - 
 I separate, and aytipw = I unite. 
 
THE BIETH OF SCIENTIFIC CHEMISTRY 21 
 
 endeavour to evince their salt sulphur and mercury to 
 be the true principles of things." In this work Robert 
 Boyle (1626-1691) strongly contests "the doctrines of 
 the four elements, and the three chymical principles 
 of mixed bodies." From the quotation just made, as 
 well as from the remarks of the debaters, it is clear 
 that a confusion of ideas had taken place with regard 
 to the Aristotelian philosophy, and that actual earth, 
 fire, air, and water, had come to be assumed by many 
 as being actually present in, and as forming real con- 
 stituents of diverse forms of matter there had been 
 in fact a resuscitation of the original Empedoclean 
 views. 
 
 Boyle, in language which knows nothing of the 
 mystic and rococo style of alchemical literature, up- 
 holds the modern view of elements, defining them as 
 certain primitive and simple bodies which, not being 
 made of other bodies or of one another, are the in- 
 gredients of which all those called perfectly mixed 
 bodies are immediately compounded, and into which 
 they are ultimately resolved. 
 
 It is clear that if gold and silver are elements in 
 this sense, i.e., simple undecomposable, indestructible 
 forms of matter created once for all in fixed and un- 
 alterable quantities by a special act of creation, then 
 all attempts at transmutation are necessarily in vain. 
 The steady growth, since Boyle's day, of the conviction 
 that gold and silver, and indeed all the metals, are 
 elements in this sense, has been contemporaneous with 
 the gradual recognition by the public of the futility 
 
22 LAW AND THEORY IN CHEMISTRY 
 
 of alchemistic aims, and of the deceptions practised 
 by the pseudo- alchemists. Yet there still exist a few 
 individuals who, rejecting the results and opinions of 
 others, and doing their own thinking, entertain firm 
 beliefs in such chimeras as the perpetual motion, the 
 quadrature of the circle, 1 and the philosopher's stone. 2 
 
 Boyle maintains that it is impossible to state offhand 
 the number of the elements. He further postulates 
 a corpuscular structure of matter, and conceives of 
 chemical combination as an approximation of corpuscles 
 of different natures ; of decomposition as the result of 
 the presence of a third kind of corpuscle capable of 
 exerting on one of the combined corpuscles a greater 
 attraction than that exerted by the corpuscles already 
 combined with it. Heretofore the ideas entertained 
 respecting chemical combination had in general been 
 very fanciful. Although the conclusion was not in 
 any way implicated in the Aristotelian conception of 
 elementary principles which held sway, yet the forma- 
 tion of a new substance had been regarded as a special 
 creation following on a real destruction of the combin- 
 ing bodies. 
 
 Boyle was the first to distinctly state that the 
 peculiar properties of substances disappear on the 
 occurrence of chemical combination or rather become 
 merged into the new and unforeseen properties of the 
 
 1 " Quadrature of the circle found. Lessons given. Fee five 
 shillings." Extract from advertisement appearing in Nature, August 
 17, 1893. 
 
 2 A good picture of this type of man is given in Balzac's L'Alchi- 
 miste. See also Dumas, Memoirs of a Physician. 
 
THE BIRTH OF SCIENTIFIC CHEMISTRY 23 
 
 compounds formed. In this connection he drew a 
 hard and fast line between mixtures and true chemical 
 combinations a distinction which will be entered into 
 in more detail in Chapter III. 
 
 Boyle further advocated the pursuance of the science 
 of chemistry independently of utilitarian ends. He 
 combated the view that it was the handmaid of this 
 or that science, it was, according to him, a self-con- 
 tained and independent part of the great study of 
 nature. In the mere advancement of knowledge for 
 its own sake he found a sufficient spur to a devoted 
 study of chemistry. 
 
 In the next chapter we shall see how Boyle's ideals 
 were realised ; how the phenomenon of combustion 
 was disinterestedly studied for the sake of the light 
 which it was hoped would by this means be shed on 
 chemical theory generally ; and how the adoption and 
 elaboration of Boyle's conception of the elements was 
 one of the chief factors in Lavoisier's happy co-ordina- 
 tion of the apparently isolated and vaguely expressed 
 items of knowledge of his day into a harmonious and 
 perspicuous whole. 1 
 
 1 The reader, who may be incited by the short sketch of the history 
 of Alchemy given in the text to a deeper study of the subject, is 
 referred to the following literature : 
 
 Article " Alchemy," Enc. Brit. ; Rodwell, "Birth of Chemistry;" 
 Meyer, "History of Chemistry ;" Thomson, "History of Chemistry ;" 
 Kopp, "Geschichte der Chemie," and " Die Alchemic in alterer und 
 neurerer Zeit ; " Berthelot, " Les origines de L'Alchimie ; " Ben 
 Jonson, "The Alchemist." (Morley's Universal Library.) 
 
CHAPTER II. 
 
 THE PHLOGISTIC PERIOD AND THE BEGINNINGS 
 OF CHEMICAL THEORY. 
 
 THE phenomenon of fire is such an important factor 
 in chemical change, both as an agent and as a result, 
 that the chemists of the seventeenth and eighteenth 
 centuries regarded it as the essential phenomenon of 
 chemistry. To found a consistent and competent 
 theory of combustion was in their opinion almost 
 tantamount to the founding of a satisfactory theory 
 of general chemistry ; and in the light of modern 
 science, which teaches that so many important chemical 
 reactions are either changes of the nature of a com- 
 bustion (i.e., are oxidations) or changes of the nature 
 of the reversal of combustion (i.e., reductions), this 
 opinion finds some justification. 
 
 The first theory of chemistry instituted by Becher 
 (1635-1682), and developed by Stahl (1660-1734), 
 was essentially a theory of combustion, and is nothing 
 else than a special development in one^ direction of 
 Becher's modification of the alchemistic views on the 
 nature of matter. 
 
 According to Becher the fundamental constituents 
 of inorganic matter were not sulphur, mercury, and 
 
THE PHLOGISTIC PEKIOD 25 
 
 salt, but the three earths, the mercurial or fluid earth, 
 the vitreous or fusible earth, and the combustible or 
 fatty earth. The latter was called terra pinguis. 
 
 Stahl especially elaborated this idea of a terra 
 pinguis. According to him all combustible bodies (sul- 
 phur, phosphorus, carbon, metals, &c.) were compounds 
 containing as an essential ingredient a fiery principle 
 which he renamed "phlogiston." In the process of 
 burning this phlogiston escaped from combination 
 often in such quantities and with such intensity as to 
 produce the phenomenon of flame. 1 In the case of 
 the combustible metals, the residue or earthy powder 
 remaining after the phlogiston had escaped was termed 
 a calx ; hence such metals were regarded as com- 
 pounds of calx with phlogiston. In the case of what 
 are now generally but loosely known as the non- 
 metals, acids resulted from combustion, hence these 
 were regarded as compounds of acids and phlogiston. 
 It was not at first noticed that the calx or acid, as 
 the case might be, weighed more than the substances 
 from which they had been produced chemistry was 
 as yet in its purely qualitative stage. 2 
 
 When the calx or acid was reheated with a body 
 rich in phlogiston, such as charcoal, it combined with 
 
 1 Death by " spontaneous combustion," which was firmly believed 
 in during the eighteenth century, and has even figured in the fiction 
 of the nineteenth (see Bleak House, Preface and chapter xxxii.), 
 met with a very simple "explanation" in terms of the phlogiston 
 theory. 
 
 2 Boyle seems to have been the first to notice this increase in weight. 
 He ascribed it to the combination of the burning body with " igneous 
 particles." 
 
26 LAW AND THEORY IN CHEMISTRY 
 
 the phlogiston of the charcoal, reforming the burnt 
 substance. Such was the phlogistonist's description 
 of reduction. But other bodies rich in phlogiston, 
 such as sulphur, flour, sugar, effected the transforma- 
 tion of a calx or acid into identically the same metal 
 or non-metal as did charcoal. 1 Hence it was argued 
 that there was but one kind of phlogiston. 
 
 This phlogiston was generally regarded as a definite 
 material substance. Stahl looked forward to its isola- 
 tion, and seems to have expected a solid earthy body 
 insoluble in water. In fact the matter of solubility 
 appeared to be merely a question of a greater or less 
 amount of combined phlogiston. Phosphorus and 
 sulphur, bodies rich in phlogiston, are insoluble in 
 water, but the acids produced from these substances 
 by the escape of phlogiston are eminently soluble. 
 The quantity of phlogiston in a substance was in fact 
 believed to condition not only its solubility but all 
 its properties its activity or inertness, its stability 
 or instability, its acidity or basicity, &c. 
 
 Although Stahl seems to have been in favour of 
 a solid phlogiston, others regarded it as gaseous, and 
 indeed went so far as to identify it with hydrogen. 
 When metals dissolve in acids a gas escapes, which 
 if collected and heated with the calx recoverable from 
 the solution the calx in this instance really being 
 
 1 In very early times it was known that wheat has the power of 
 revivifying a metal from its calx or ashes, and it is said that it was 
 partly on account of this property that wheat was made the popular 
 symbol of the resurrection. 
 
THE PHLOGISTIC PERIOD 27 
 
 a salt restores the metal. But metal is calx and 
 phlogiston ; the gas disengaged on dissolving metals 
 in acids is therefore phlogiston. 
 
 The doctrine of phlogiston as just sketched, after 
 reigning triumphantly for some half century, finally 
 succumbed to the increasing knowledge of the con- 
 stitution and nature of the atmosphere. This know- 
 ledge shed quite a new and certain light on the 
 phenomena of combustion, and was co-ordinated by 
 Lavoisier into the theory of combustion held at the 
 present day. 
 
 Let us now trace the chief stages in the growth of 
 our knowledge of the atmosphere. 
 
 Eobert Hooke (1635-1703), a contemporary of 
 Becher, and sometime assistant to Boyle, published in 
 1665 his conception of combustion in his Micro- 
 grapliia. He believed that there existed in air a 
 fractional quantity of the same kind of gas as is ob- 
 tained by heating saltpetre, or nitre, and that com- 
 bustion consisted in the solution of the combustible 
 body by this gas. Hooke pointed out many analogies 
 between combustion and the solution of solids in liquid 
 menstrua. 1 
 
 Mayow (16451679), an Oxford physician, worked 
 out in more practical detail the ideas of his contem- 
 porary Hooke. He showed that when a metal burns 
 in air, the volume of air is actually lessened, and that 
 
 1 It is interesting to note that Hooke was on the point of anticipating 
 Rumford. He clearly stated that fire or flame is not an element, but 
 a phenomenon resulting from the agitation of particles. 
 
28 LAW AND THEORY IN CHEMISTRY 
 
 the calx residue weighs more than the original metal. 
 This he explained by suggesting that the metal had 
 combined with the particles of the nitre air, as he 
 called it, present in common air, leaving a residue of 
 inactive air. The same fact was glossed over by the 
 phlogistonists, in terms of the assertion that phlogiston 
 was a principle of levity, and that therefore its escape 
 from a body during combustion rendered the body 
 heavier. And so supremely popular was the phlogis- 
 ton theory at the time, that the reasonableness and 
 common sense of Mayow's explanation had no weight 
 against the transcendental artificiality and uniqueness 
 of its own. 
 
 Mayow's experiments were so much to the point, 
 that it is hard to see how it came about that their full 
 significance was not recognised till they were practi- 
 cally rediscovered piecemeal a century afterwards. 
 
 The next advance towards an accurate knowledge of 
 combustion was Black's discovery and investigation of 
 what he called "fixed air" the gas now known as 
 carbonic acid gas, or carbonic anhydride. 
 
 The attention of the medical world was about this 
 time (1728-1799) directed to quicklime, magnesia, and 
 allied substances, in the role of efficacious remedies in 
 the treatment of stone in the bladder, and Black, being 
 greatly interested in medicine, undertook a chemical 
 investigation of these new remedies. Up to this time 
 the carbonates of lime, magnesia, and the alkalis had 
 been regarded as elements, or rather simple bodies, 
 which when heated formed quicklime, caustic alkali, 
 
THE PHLOGISTIC PERIOD 29 
 
 &c., in virtue of their combination with the phlogiston 
 escaping from the burning coal. 
 
 Black chiefly investigated what was then called mild 
 magnesia, i.e., magnesium carbonate. He first proved 
 that when the carbonate is heated, it loses weight. 
 This loss is not due to a combination with phlogiston 
 of negative specific gravity, but to the expulsion of a 
 gas which he called fixed air, because it could be again 
 refixed in a sort of latent condition by the resulting 
 caustic magnesia. He showed that the same gas is 
 produced when chalk is calcined or treated with acids, 
 that it exists in the breath of animals, and is evolved 
 in large quantities during fermentation. Further, that 
 animals die when placed in it, and that it does not 
 support combustion. Nothing regarding the composi- 
 tion of this gas was, however, hinted at. 1 
 
 The next step towards the full light was taken by 
 the English divine Priestley (1733-1804). Priestley 
 did not allow his Calvinistic doctrines to usurp his 
 whole interest and leisure, but he gave much attention 
 to chemistry, making a speciality of that branch of it 
 dealing with the different kinds of air or as we should 
 say now, the different kinds of gases. 
 
 While taking charge of a chapel in Leeds, he hap- 
 pened to live near a brewery, in which he spent much 
 time studying the properties of fixed air, which Black 
 had shown to be produced in large quantities during 
 fermentation. On being compelled by circumstances 
 
 1 Black's paper has been republished in convenient pamphlet form 
 by the Alembic Club (Sampson & Co.). 
 
30 LAW AND THEORY IN CHEMISTRY 
 
 to move from the purlieu of this laboratory, he did not 
 drop his interest in fixed air, but began to make it 
 artificially from chalk as taught by Black. In order to 
 store and examine gases, he made extended use of and 
 popularised Mayow's method of collecting them over 
 water in what he called a pneumatic trough, a method 
 still in vogue. He soon noticed that the fixed air 
 dissolved to a considerable extent in water, and that 
 the solution had medicinal properties. From this dis- 
 covery dates the manufacture of artificial mineral 
 waters. 
 
 Priestley, having fairly exhaustively worked out the 
 properties of fixed air, turned his attention to new airs. 
 In this investigation he seems to have been guided by 
 no principle of selection, but at haphazard he subjected 
 to the action of the sun's rays concentrated by a lens 
 any compound that he happened to think of or chance 
 upon; the gas, if any, produced being collected over 
 mercury. 1 One of the substances thus experimented on 
 was the so-called red precipitate. From this, in 1772, 
 he obtained a gas in which inflammable bodies burned 
 vigorously, and "which had all the properties of com- 
 mon air, only in much greater perfection." He gave it 
 the name of dephlogisticated air, and regarded it merely 
 as common air quite free of impurities and admixtures. 
 He did not regard it as a constituent of air as Hooke 
 and Mayow some years previously had done with truer 
 insight. Air was to Priestley essentially a simple sub- 
 
 1 A statue of Priestley in Birmingham represents him conducting 
 this operation. 
 
THE PHLOGISTIC PERIOD 31 
 
 stance, and his dephlogisticated air was merely common 
 air in a state of great purity. In 1774, being in Paris, 
 he demonstrated to Lavoisier his method of making 
 dephlogisticated air. Some years later, as we shall see, 
 Lavoisier made great use of this knowledge, and re- 
 christened dephlogisticated air, giving it its present 
 name, oxygen. 
 
 Priestley was a confirmed phlogistonist. The follow- 
 ing is something like his idea of what takes place dur- 
 ing combustion. When a substance burns, phlogiston 
 escapes from it into the air, which is invested with a 
 great affinity or longing for this principle of fire. Sub- 
 stances cannot burn out of contact with the air, or other 
 substances, such as nitre, which possess this strong 
 attraction for phlogiston. But in proportion as the air 
 round a burning body becomes more and more charged 
 with phlogiston, the poorer a supporter of combustion 
 does it become, till finally it may become, so phlogis- 
 ticated so noxious, as he expressed it that it will 
 actually quench a flame immersed in it. The air 
 obtained from red precipitate was virgin air untainted 
 by any phlogiston, and so energetically did it long for 
 this principle of fire that even at ordinary temperatures 
 it drew it away from several substances, thus causing 
 them to tarnish. Iron nails, as we know, soon tarnish, 
 i.e., rust, when immersed in impure oxygen. 
 
 If carbon is heated with metallic calx it gives up the 
 phlogiston in which it is rich to the calx, reforming 
 the pure metal from it. But according to Priestley 
 the air that has surrounded a burning body is rich in 
 
32 LAW AND THEOEY IN CHEMISTRY 
 
 phlogiston ; it then, by a parity of reasoning, ought to 
 reduce metals from their calces. If Priestley had but 
 tried the experiment here suggested, he would have 
 met with such a positive failure that one can hardly 
 conceive how he could have continued to cherish his 
 views on the nature of combustion. 1 It will be noticed 
 that Priestley did not attempt to explain why the air 
 diminished when combustion took place in a confined 
 space. 
 
 Priestley, essentially a preparateur, also experimented 
 in his own casual way on nitrous air (nitric oxide), 
 vitriolic acid air (sulphur dioxide), muriatic acid air 
 (hydrochloric acid), alkaline air (ammonia), and inflam- 
 mable air (hydrogen), which latter had been discovered 
 by Cavendish in 1766 and 'recognised by him as an 
 individual gas. 
 
 In his experiments on inflammable air, Priestley very 
 nearly anticipated Cavendish's discovery made three or 
 four years afterwards, of the composition of water. 
 Indeed, Priestley's mind was so saturated with the 
 importance of the chimera phlogiston, 2 that for once 
 at least in his life he failed of a customary good 
 fortune to which he alludes in the .following words : 
 " In looking for one thing I have generally found 
 another, and sometimes a thing of much greater im- 
 portance than that which I was in quest of." 
 
 1 It is easy to point out many similar inconsistencies in the theory 
 of phlogiston. For instance, the ash from burnt charcoal ought to 
 weigh more than the original charcoal. 
 
 2 Priestley's last work is entitled The Theory of Phlogiston 
 Established. 
 
THE PHLOGISTIC PERIOD 33 
 
 In experimenting with the various airs, Priestley 
 always tried their effects on animal and vegetable life. 
 For this purpose mice and sprigs of mint seem chiefly 
 to have been victimised. In this connection it is 
 interesting to recall that he distinctly noted that won- 
 derful piece of natural economy whereby the gaseous 
 waste-products of animals serve as the food of plants, 
 and the atmosphere is maintained in a state of fresh- 
 ness and purity compatible with the prolonged existence 
 of life. 
 
 It is a great source of regret to read that Priestley, 
 who by his blind and undirected enthusiasm for dis- 
 covery laid some of the foundation-stones of modern 
 chemistry, was so little appreciated by a large section 
 of his countrymen that he finally left his native land 
 with the remark, "When the time for reflection shall 
 come, my countrymen will, I am confident, do me more 
 justice." 
 
 Before we pass on to the brilliant co-ordination of 
 the facts of combustion and their true explanation by 
 Lavoisier, we must mention briefly the work of that 
 peculiarly ascetic philosopher, the Hon. Henry Caven- 
 dish work which furnished Lavoisier with some of his 
 most weighty data. 
 
 In 1766 Cavendish discovered that when about two 
 volumes of inflammable air are exploded with one volume 
 of dephlogisticated air, water is produced. Further, 
 if ordinary atmospheric air is used instead of dephlo- 
 gisticated air, water is still produced, but a residue 
 of phlogisticated air, i.e., what we now call nitrogen, 
 
34 LAW AND THEORY IN CHEMISTRY 
 
 always remains. Hence Cavendish argued that air was 
 not a simple individual substance, but a mixture of phlo- 
 gisticated and dephlogisticated airs. It would seem, 
 however, that Cavendish did not clearly grasp what 
 seems to us such an evident conclusion from his experi- 
 ment, viz., that water is a compound of inflammable 
 and dephlogisticated airs. 1 He had been brought up 
 on the phlogiston doctrine, and could never thoroughly 
 free himself from its trammels. Although he saw good 
 reasons for supposing that when a metal burns in the 
 air it combines with the dephlogisticated air to form 
 a calx leaving the phlogisticated air, yet he could not 
 find it in his heart to give support to such a heresy. 
 He therefore concluded with Priestley that the metal 
 phlogisticated the residual air, but slightly differed 
 from that scientist in admitting that some of the air 
 in a dephlogisticated state combines with the metal. 
 Combustion was a case of a partial combination associ- 
 ated with and attended by a contamination ; not of a 
 total combination attended by an elimination. 2 
 
 It is remarkable that the most fearful reign of social 
 anarchy and bloodshed the modern world has known 
 should have bequeathed to us through Lavoisier so 
 unexpected a gift as a remodelled and thoroughly 
 scientific theory of chemistry. Lavoisier was born in 
 Paris in 1743. His peaceful life of scientific research, 
 passing amidst all the horrors culminating in the reign 
 of terror, was finally ended on the guillotine. 
 
 1 James Watt seems to have been the first to arrive at this conclusion. 
 
 2 The Alembic Club has reprinted in convenient pamphlet form some 
 of the more important of Cavendish's papers. 
 
THE PHLOGISTIC PERIOD 35 
 
 About the year 1770 Lavoisier turned his attention 
 to combustion. His great success in this field of study 
 must be largely attributed to the importance he placed 
 on the incessant use of the balance as an instrument of 
 research; an instrument with which he, early in his 
 scientific career, established the principle of the con- 
 servation of mass l a principle lying at the foundation 
 of the science of to-day, and familiar to all. In no 
 chemical change is matter either created or destroyed; 
 the sum of the masses of the factors of a chemical 
 change being identically equal to the sum of the masses 
 of the products. The form of matter can le changed, 
 lut not its quantity. This principle, which strikes the 
 materialistic mind of to-day so much in the character of 
 a truism, involved quite a revolution in the settled views 
 of the end of the eighteenth century. As has before 
 been stated, the production of new forms of matter 
 was regarded as an act of special creation, having no 
 quantitative relation to the destruction or annihilation 
 which necessarily preceded it. In Lavoisier's time, too, 
 heat or caloric was popularly regarded as material, and 
 consequently its greater or less rdle in any chemical 
 change would presumably more or less affect the mass 
 relations. But Lavoisier proved that the mass of a 
 compound AB is, within the limits of experimental 
 error, exactly equal to the mass of A and the mass of 
 B combining to form that compound, and is quite 
 independent of the heat evolved or absorbed during 
 the combination. 
 
 1 This principle was assumed in many of the ancient philosophies. 
 
36 LAW AND THEORY IN CHEMISTRY 
 
 Although Black and Cavendish had done accurate 
 pieces of quantitative work, yet it is with Lavoisier that 
 chemistry essentially passed from the qualitative to the 
 quantitative stage. It was he who ever insisted on the 
 paramount importance of the question, "How much?" 
 in every chemical inquiry. 
 
 In connection with combustion, Lavoisier first con- 
 firmed the fact that when a metal burns it increases in 
 weight, and then proceeded to prove in the case of tin 
 that the increase in weight is due to air absorbed, and 
 is equal to the weight of the absorbed air. He noticed 
 that it was not the air as a whole that was absorbed, 
 but only a constituent thereof, the unabsorbed con- 
 stituent being different from common or fixed airs. 
 From classical experiments on red precipitate (HgO), 
 suggested by Priestley, he concluded that the con- 
 stituent absorbed by burning bodies was nothing else 
 than Priestley's dephlogisticated air. 
 
 While this air can be expelled from, the red calx of 
 mercury by mere heating a point whose explanation 
 had always baffled the phlogistonists the calces of 
 other metals had to be heated with carbon, and the air 
 disengaged under these conditions Lavoisier proved to 
 be identical with Black's fixed air, which therefore must 
 consist of carbon and dephlogisticated air, and which he 
 therefore renamed carbonic acid gas. The carbon com- 
 bined with the dephlogisticated air of the calx, just as 
 it did with that of the atmosphere when burned therein. 
 This same carbonic acid gas Lavoisier showed to be 
 formed during the respiration of animals, thus trans- 
 
THE PHLOGISTIC PERIOD 37 
 
 forming a suspicion which had arisen in the mind of 
 Paracelsus into a certainty. Mayow a century pre- 
 viously had come to the same conclusion, though not 
 in so definite a manner. 
 
 Since the combination of dephlogisticated air with 
 the majority of substances investigated by Lavoisier 
 produced acids, he renamed the gas oxygen, or the acid 
 producer; a nomenclature which modern research has 
 shown to be not altogether appropriate. 
 
 On burning phosphorus in a confined volume of air, 
 Lavoisier found that after about i of the air had dis- 
 appeared the phosphorus was extinguished, and all 
 other combustible substances refused to burn in the 
 residual gas. He regarded this noxious air as a true 
 constituent of common air actually present therein 
 before the burning, and called it first "moufette atmos- 
 pherique," and then later " azote," because animals died 
 when immersed in it. Rutherford in Edinburgh simul- 
 taneously isolated this gas by treating with caustic 
 potash solution air in which animals had been confined 
 for some time. He gave it the more popular name 
 nitrogen. From his experiment with burning phos- 
 phorus, Lavoisier rightly concluded that air was a mix- 
 ture of about i oxygen and 4 nitrogen or azote. 
 
 Having given a simple explanation of calcination or 
 oxidation and reduction by carbon, Lavoisier now turned 
 his attention to the process of reduction, i.e., the re- 
 forming of the metals from their calces by means of 
 hydrogen. To this end he repeated Cavendish's experi- 
 ment on the composition of water, rightly interpreting 
 
38 LAW AND THEOEY IN CHEMISTRY 
 
 the results and completing the proof by passing steam 
 over heated iron, thus obtaining a calx and inflammable 
 air which he thenceforward called hydrogen. When a 
 calx is heated with hydrogen, the metal reappears, not 
 because the calx takes up the phlogiston, in which the 
 hydrogen is rich, but because the hydrogen wrests the 
 oxygen from the calx and forms therewith invisible 
 steam. 
 
 Lavoisier regarded the solution of a metal in a dilute 
 acid as a reaction taking place in two stages. First, 
 the metal decomposed the water, setting free the 
 hydrogen, and combining with the oxygen to form a 
 calx ; and then this calx combined with the acid, form- 
 ing a salt soluble in the water holding the acid in 
 solution. 
 
 Here truly was a complete renversement of current 
 ideas. To the phlogistonist the metal or combustible 
 was more complex than the calx or product to which 
 it gave rise. The metal or combustible consisted of 
 matter plus phlogiston, the calx or product was metal 
 or combustible minus phlogiston. To the Lavoisierian 
 or antiphlogistic school, however, the calx was more 
 complex than the metal it was a compound, while the 
 metal was an element. 
 
 When we look back on Lavoisier's life-work, we see 
 it to consist of co-ordination and generalisation rather 
 than discovery. He made use of the facts brought him 
 by what Huxley has called the hod-carriers of science. 
 In making use of these contributions he sometimes, 
 unfortunately for his moral reputation, forgot their 
 
THE PHLOGISTIC PERIOD 39 
 
 sources, and claimed them as original. Lavoisier's 
 attitude towards Priestley, Cavendish, and others, whose 
 data he used, is still considered a subject for severe 
 criticism. 
 
 Undoubtedly a great factor in Lavoisier's success 
 was his unconditional rejection of the old, vague, ele- 
 mentary principles of the alchemistic era, and his 
 adoption of the conception of elements and compounds 
 as developed by Boyle, and now generally received. 
 
 The theory of phlogiston was doomed. Lavoisier's 
 explanations steadily gained in favour, and the year 
 _1_785 marked the complete ascendency of his anti- 
 phlogistic views. In France his teaching was received 
 partly from intellectual conviction, partly from a feel- 
 ing of patriotism, stimulated by the title La Chemie 
 Frangaise, which Fourcroy was presumptuously pleased 
 to bestow on his countryman's theory. 
 
 The majority of British and German chemists also 
 expressed their allegiance to the new theory before its 
 illustrious founder prematurely passed v away. 
 
CHAPTER III. 
 
 CHEMICAL CLASSIFICATION. MIXTURES, 
 COMPOUNDS, ELEMENTS. 
 
 THE fundamental classification of chemistry is com- 
 prised in the two expressions, homogeneous or pure 
 bodies, and non-homogeneous bodies or mixtures. 
 Everything material necessarily falls into one or other 
 of these two classes. 
 
 It is sometimes quite erroneously stated that only 
 the homogeneous substances belong to the domain 
 of chemistry. This statement probably originates in 
 the fact that all the fundamental chemical laws have 
 reference to, and are only true for, homogeneous bodies ; 
 but homogeneity is the exception, not the rule in 
 nature, 1 and chemistry is a natural science. Matter 
 invariably comes into the laboratory in a state of 
 non-homogeneity or mixture, and it is always the 
 first and often the most difficult task of the chemist 
 to resolve such a mixture into its homogeneous con- 
 stituents. 
 
 1 Witness the derivation of the word metal. According to Pliny 
 this word is derived from /uer' &\\a signifying together with other 
 tilings. There is, however, another possible derivation of the word 
 metal, viz., from ^TO\\OV = a mine, connected with ^eraXXciw = 7 
 search for diligently. 
 
CHEMICAL CLASSIFICATION 41 
 
 The study of coal tar, for example, has resulted in 
 most important advances in the theory of chemistry. 
 Yet the study of coal tar as a whole would have been 
 quite barren of valuable results. Tar being a variable 
 mixture, the investigation of its properties would have 
 led to inconstant results, and made impossible any 
 attempts at productive generalisation. But by the 
 process of fractional distillation (vide infra), coal tar 
 has been sorted into a number of homogeneous con- 
 stituents, and it is the study of these constituents 
 individually, not of tar as a whole, that has profitably 
 engaged the attention of chemists. 1 Thus benzene 
 
 1 The following scheme represents the approximate percentages 
 obtained, and the stages passed through in the separation (see post, 
 p. 47) of six of the more important homogeneous constituents of coal 
 tar : 
 
 Up to 
 First Ei 
 
 I 
 
 Water. 
 
 110 C. 110 to 
 innings. Light 
 
 ? / 
 
 t\ / 
 
 A \ First 
 ^ \ portion. 
 
 Later j 
 First was 
 acid anc 
 thenmixe 
 
 220 C. 210 to 
 Oil. Carbo 
 
 / 
 
 \y 
 v 1 
 ^Carbolic 
 / Acid 
 
 240 C. 240 to 270 C. 
 lie Oil. Creosote Oil 
 or 
 Heavy Oil. 
 
 Naphtha- 
 lene 
 
 270-* 
 Anthracene 
 Oil. 
 
 Anthracene 
 
 to per ct. 
 
 / 1 per ct. 
 Residue. 
 
 ortiou. 
 lied with 
 I alkali, 
 d with A. : 
 
 10 per ct. 
 
 
 
 Benzene Toluene 
 
 Xylenes Residue 
 and other (burning 
 t. hydro- Naphtha), 
 carbons 
 
 b. p. 140 to 150. 
 Solvent Naphtha. 
 
 b. p. 81 b. p. Ill 
 1 per cent. <lpercer 
 
 35 per cent. 
 
42 LAW AND THEORY IN CHEMISTRY 
 
 is a definite substance with fixed and invariable pro- 
 perties, and its constants determined for a preparation 
 from any one coal tar will hold good for a specimen 
 obtained from any other tar. 
 
 What, we may now inquire, is the distinguishing 
 feature of the class of non-homogeneous bodies ? Briefly 
 stated it is this ; a non-homogeneous substance or mix- 
 ture can be resolved into unlike portions by processes 
 of mere sorting, and the sum of the energies of the 
 sorted portions is equal to the energy of the mixture 
 from which they were sorted. 1 
 
 If one simply mixes a red powder A with a green 
 powder B (no heat evolution or absorption accompany- 
 ing the process) the mixture can be resolved into its 
 constituents again by patient "hand sorting" under 
 the microscope. 2 If flour and sugar be mixed, they 
 can be separated again by treating the mixture with 
 water, in which the sugar alone is soluble. The 
 mere mixture of flour and sugar does not alter the 
 properties peculiar to each of these substances in the 
 pure state. 
 
 Even when all the constituents of a mixture are 
 soluble in a given medium, they may often be com- 
 
 1 Mixtures or emulsions of such liquids as are ordinarily spoken of 
 as immiscible, e.g., oil and water, are only temporary. Such mixtures 
 possess more energy than the separated constituents ; but this surplus 
 of energy, depending on the area of the surfaces of separation of the 
 different liquids, sooner or later runs down into heat, and contem- 
 poraneously the mingled liquids sort themselves into easily separable 
 layers. [See Maxwell, Theory of Neat, chap, xx.] 
 
 2 Dextro- and laevo-tartaric acids were first separated from each 
 other and obtained pure by this method. 
 
CHEMICAL CLASSIFICATION 43 
 
 pletely separated by repeated crystallisation from that 
 medium. Thus when an aqueous solution of potassium 
 chloride and potassium chlorate is evaporated, the less 
 soluble chlorate crystallises out first, and afterwards 
 the more soluble chloride. This method of separation 
 is actually adopted in the commercial preparation of 
 chlorate of potassium. At Stassfiirt a mineral com- 
 posed of potassium and sodium chlorides is found in 
 quantity. In order to separate the constituents, a cold 
 saturated solution is first made of the mixture. This 
 solution is then raised to the boiling point. In pro- 
 portion as the water evaporates the sodium chloride 
 crystallises out. After a certain time the solution is 
 allowed to cool, when potassium chloride separates out. 
 For this salt, while much more soluble at high tempera- 
 tures than common salt, decreases in solubility very 
 rapidly indeed as the temperature is lowered. By 
 alternately repeating the boiling and cooling, large 
 quantities of the chlorides are separated. 
 
 Again, oxygen and nitrogen are both soluble in 
 water, but oxygen to a greater degree than nitrogen. 
 Common air is a mixture of these two gases, conse- 
 quently the "air" dissolved by water (which "air" 
 is disengaged again on boiling the water) is much 
 richer in oxygen than common air. If the process 
 of treating water with air under pressure, and then 
 boiling the resulting solution, were repeated often 
 enough fairly pure oxygen could be eventually sorted 
 out. An investigation of the results of a single opera- 
 tion of the above process will be instructive and will 
 
44 LAW AND THEORY IN CHEMISTRY 
 
 afford an opportunity of illustrating important laws, 
 one discovered by Dalton. and the other by Henry and 
 Dal ton. The data requisite for this investigation are 
 as follows : 
 
 Per cent. 
 
 Composition of air by volume, oxygen . . 20 '9 
 nitrogen . . 79'1 
 
 Approximate coefficient of absorption of oxygen . '04 
 nitrogen '02 
 
 The coefficient of absorption of a gas denotes that 
 volume thereof (measured at C. and under a pressure 
 of 760 mm. of mercury) which is dissolved by one 
 volume of water when an excess of the gas is presented 
 to the water under a pressure of 760 mm. of mercury. 
 Thus one cubic foot of water dissolves '04 of a cubic foot 
 of oxygen (measured at C. and 760 mm.) when an 
 excess of the gas is allowed to stand over the water 
 at normal pressure. 
 
 Dalton' s approximative law l of partial pressures 
 (1802) states that the total pressure of a mixture of 
 gases is the sum of the partial pressures exerted by 
 the constituents of the mixture in the space occupied 
 by the mixture. Suppose the air exerts a pressure 
 of 760 mm. Then, from Boyle's law and the above 
 data, the partial pressure of the oxygen P must be 
 given by the equation 
 
 P x 100 - 20-9 x 760. 
 
 P = 760 x -209 mm. 
 
 Similarly the partial pressure of the nitrogen P n is 
 
 1 "An approximative law expresses only a portion of a complex 
 phenomenon the limit towards which the phenomenon aims." 
 
CHEMICAL CLASSIFICATION 45 
 
 equal to 760 x '791 mm. In conformity with Dalton's 
 law, it will be noticed that P + P n = 760 mm. 
 
 Henry and Dalton's law (1803-1807) is an approxi- 
 mative law which states that the masses of the less 
 soluble gases dissolved at ordinary temperatures are 
 proportional to the partial pressures exerted by these 
 gases. 1 Let d be the density of oxygen at and 
 760 mm. Then one cubic centimetre of water dissolves 
 -Q4:d grams of oxygen when the pressure is 760 mm. 
 By Henry and Dalton's law it will therefore dissolve 
 
 x 760 x -209 
 
 - or -Old x -209 = -00836 x d grams, 
 
 760 
 
 when the pressure is 760 x '209 mm. The volume of 
 this oxygen is "00836 cc. Similarly -02 x '791 cc. of 
 nitrogen are absorbed by one cubic centimetre of 
 water. Hence the composition of the dissolved air 
 is approximately 
 
 Oxygen . . . . 34*5 per cent. 
 Nitrogen . . . . 65 '5 per cent. 
 
 In other words, whereas air contains only about 21 
 per cent, of oxygen, the dissolved air contains about 
 34J per cent. 2 
 
 1 It is frequently stated that the law of Henry and Dalton furnishes 
 a criterion as to whether a gas is physically or chemically dissolved. 
 If the law is obeyed, then it is said to be a case of physical absorption, 
 otherwise it is chemical change. Such a hard and fast distinction 
 cannot be maintained. Solution is a complex phenomenon, and even 
 in the simplest cases it seems almost certain that true chemical action 
 and simple physical change are inextricably involved (see p. 55). 
 Further, it is probable that at low enough temperatures no gas would 
 be found to obey Henry and Dalton's law. 
 
 2 If the process indicated in the text were to be repeated with this 
 
46 LAW AND THEORY IN CHEMISTRY 
 
 Many mixtures may be resolved by taking advantage 
 of differences in the volatilities of their constituents. 
 When in the analysis of an inorganic substance by 
 the usual methods we come to test for potassium, it 
 is necessary before performing the test to remove 
 the salts of ammonium which have been added earlier 
 in the course of analysis. This is effected by eva- 
 porating the solution to dryness and then heating 
 the residue to low redness, when the more volatile 
 ammoniacal salts will pass off, leaving behind the 
 less volatile potassium salts. 
 
 A mixture of liquids can often be sorted by frac- 
 tional distillation ; that is, by taking advantage of 
 differences in the boiling point of its constituents. 
 If a mixture of ether (B.P. 35) and aniline (B.P. 
 181) be heated in a suitable apparatus, the tempera- 
 ture of the mixture rises quickly till the neighbour- 
 hood of the boiling point of ether is reached, when 
 it remains fairly constant till nearly all the ether is 
 distilled off. The temperature then rises rapidly to 
 the boiling point of aniline, when nearly pure aniline 
 is obtained as a distillate. It is not in general pos- 
 sible to effect in this way a complete separation of 
 
 air of 34| per cent, oxygen, the composition of the air resulting from a 
 second solution in water would be approximately 
 
 47*5 Oxygen, 
 52-5 Nitrogen, 
 
 and a third solution in water would make the composition of the dis- 
 solved air approximately 
 
 75 Oxygen, 
 
 25 Nitrogen. 
 
CHEMICAL CLASSIFICATION 47 
 
 mixed liquids in one operation, it is only when the 
 process has been repeated systematically several times 
 that perfect separation is attained. 
 
 This process of fractional distillation is the first and 
 most important stage in the resolution of coal tar 
 into its homogeneous constituents (see p. 41) ; the final 
 isolation of these constituents in a state of complete 
 purity being effected by other sorting processes adapted 
 to the individual cases. 
 
 As will readily be conceived, a mixture of liquids 
 possessing the same or nearly the same boiling points 
 does not yield to this method of fractional distillation, 
 and indeed in many cases where the boiling points of 
 the constituents of a mixture differ fairly widely, 
 separation is found to be impracticable. Let A and 
 B be the two constituents of a liquid mixture ; then 
 it would appear that a separation of A and B by 
 fractional distillation is only possible when the boiling 
 point of every conceivable mixture of A and B lies 
 intermediate between the boiling points of the free 
 constituents A and B. This is the case with ether 
 and aniline, and therefore all mixtures of these two 
 substances can, as already described, be separated by 
 fractional distillation. But it is not the case, for 
 instance, with ethyl alcohol (B.P. 78) and water; 
 hence mixtures of alcohol and water are not completely 
 resolvable by fractional distillation in spite of the large 
 difference between the boiling points of the liquids. 1 
 
 1 A mixture of 97 per cent, ethyl alcohol with 3 per cent, water 
 has a higher boiling point than either pure alcohol or pure water. 
 
48 LAW AND THEORY IN CHEMISTRY 
 
 Many mixtures may be resolved by taking advantage 
 of differences in the densities of their constituents. 
 This principle finds application in the " panning out " 
 of gold from its ores, in the sorting of diamonds 
 from their specifically lighter matrix in the so-called 
 pulsators, and in the resolution of various mixtures 
 by centrifugalisation. 1 In the analysis of a soil the 
 clay is separated from the sand by stirring up the 
 soil in water. The sand sinks to the bottom of the 
 containing vessel, while the clay remains suspended in 
 the supernatant liquid, which is poured off and filtered. 
 Such methods of resolution, wherein an inert menstruum 
 is called into play, are often grouped together under 
 the title of elutriation methods. 2 
 
 A mixture of gases of different densities can also 
 be partially resolved by a process introduced by 
 Graham (1805-1869), and called by him atmolysis 
 
 Hence pure anhydrous alcohol cannot be obtained by fractional 
 distillation. Suppose we start with a mixture of 50 parts water and 
 50 parts alcohol. Fractional distillation will resolve this into two 
 portions : (1) 48'45 parts of pure water, and (2) 51 '55 parts of 97 per 
 cent, ethyl alcohol (50 parts alcohol, and 1 '55 parts water). In other 
 words, it is fairly easy to get in a pure state some of the water mixed 
 with the alcohol, but it is impossible to get alcohol in an anhydrous 
 state from an aqueous mixture. To obtain pure anhydrous alcohol 
 recourse must be had to distillation of the 97 per cent, alcohol over 
 anhydrous baryta. 
 
 1 The centripetal force necessary to keep a particle of mass m 
 moving in a circular path of radius r with angular velocity a, is 
 numerically equal to the product mra?. Consequently the more 
 massive, or rather, the denser particles of a mixture will have a 
 greater tendency to fly tangentially off a revolving plate than the less 
 dense ones. 
 
 2 For very neat applications of elutriation methods to the proximate 
 analysis of minerals see Nature, xliii. p. 404, and xlix. p. 211. 
 
CHEMICAL CLASSIFICATION 49 
 
 a rather misleading nomenclature in this connection. 
 The lighter a gas is, the more quickly does it pass 
 through a porous partition, cceteris paribus. In the 
 annexed diagram cc represents a glass tube through 
 the axis of which runs a porous tube (e.g., the stem 
 of a churchwarden pipe) TT'. The side tube c' of the 
 glass is in connection with an exhausting pump. If 
 electrolytic gas, which is a mixture of hydrogen and 
 oxygen in the proportion of two volumes of the former 
 to one of the latter, be passed slowly through the tube 
 
 FIG. l. 
 
 TT', the pump will extract from c a mixture richer in 
 the lighter gas hydrogen than the electrolytic gas used, 
 while the mixture issuing from TT' will be richer in 
 oxygen. Finally a limit will be reached whereat the 
 repetition of the process will be unproductive of further 
 resolution. The exact statement of the connection 
 between the densities of gases and their rates of passage 
 through a thin septum of graphite, or better still, biscuit 
 ware, is known as Graham's law of diffusion. The 
 rates of diffusion of gases under the same pressure 
 vary inversely as the square roots of their densities; 
 or algebraically stated 
 
 rate of diffusion of gas x _ v density of y 
 rate of diffusion of gas y ~ J densItV oTx 
 
 D 
 
50 LAW AND THEORY IN CHEMISTRY 
 
 This law is not actually realised in practice ; it would 
 be absolutely true for "perfect gases " (i.e., ideal gases 
 which obey Boyle's law perfectly) diffusing through an 
 infinitely thin septum. 1 
 
 We may also separate some mixtures by taking ad- 
 vantage of their different diffusibilities, as in the process 
 of dialysis, also invented by Graham. When substances 
 are in solution, the rates at which they pass through 
 animal or vegetable membranes 2 differ widely. Sub- 
 stances which can crystallise pass through such mem- 
 branes readily. 3 On the other hand, solutions of bodies 
 which do not crystallise but exhibit conchoidal fracture, 
 are either quite incapable of passing, or only pass with 
 the greatest difficulty, through such membranes. The 
 former readily diffusible class of bodies Graham called 
 
 1 This law, in common with the other gaseous laws of Boyle, Charles, 
 and Avogadro (see p. 89), is a necessary consequence of the kinetic 
 theory of gases. This theory simply postulates that the pressure of a 
 gas is due to the impact on the walls of the containing vessel of the 
 small similar particles of which the bulk of the gas is made up. 
 
 Reasoning thence, the expression A / -- 2 i s arrived at by pure mathe- 
 
 \ d 
 
 matics for the average velocity of progressional movement of the 
 particles of a gas whose density is d, under a pressure of p cm of 
 mercury of density D, at a place where the acceleration due to gravita- 
 tion is g. Since the rate of diffusion of a gas varies directly as the 
 average velocity of movement of its particles, the law of Graham at 
 once follows. 
 
 2 For purposes of dialysis, a parchment made by soaking unsized 
 paper in sixty per cent, sulphuric acid for a short time, is in general use. 
 
 3 Striking exceptions to this generalisation are furnished by haemo- 
 globin, the colouring matter of the blood, and vitellin, an albumen 
 occurring largely in plants. Both of these substances, though crystal- 
 line, are non-diffusible. It has also been stated that egg albumen 
 a colloid almost as typical as glue can be obtained in the crystalline 
 state. 
 
CHEMICAL CLASSIFICATION 51 
 
 crystalloids ; to the latter non-diffusible class he gave 
 the name colloids, from the Greek word (KO\\O) for 
 glue, their typical representative. As an example of 
 the application of the method of dialysis we will take 
 the preparation of the important drug "liquor ferri 
 dialysati." Like many colloid bodies, ferric hydroxide 
 can exist in a soluble or hydrosol form, and in an in- 
 soluble or hydrogel form. A solution of the hydrosol 
 form constitutes the drug above mentioned, and it is 
 prepared as follows. The 
 hydrogel or ordinary form of 
 ferric hydroxide is soluble in 
 a solution of the crystalloid 
 ferric chloride. Such a solu- 
 tion is prepared and then 
 placed in a dialyser (Fig. 2). 
 This consists of a U tube 
 uu, made out of parchment 
 paper suspended from a rod 
 rr in a large vessel contain- 
 ing water. The crystalloidal ferric chloride diffuses 
 into the surrounding water, which should be frequently 
 changed ; while the hydrosol form of the colloidal ferric 
 hydroxide 1 remains behind in the U tube in a pure 
 state. 
 
 This process of dialysis is frequently resorted to in 
 toxicological investigations for the purpose of separating 
 the crystalloidal poisons from the colloidal contents 
 
 FIG. 2. 
 
 1 It would appear that liquor ferri dialysati is not a solution of pure 
 ferric hydroxide, but of a hydrosol ferric hydroxychloride. 
 
52 LAW AND THEOKY IN CHEMISTEY 
 
 of the stomach and intestines which would mask the 
 reactions of the poisons. 
 
 The method of refrigeration is also largely used in 
 the resolution of mixtures. Thus the impure benzene 
 obtained by fractionally distilling coal-tar is freed from 
 admixed toluene by surrounding it with ice. The 
 benzene solidifies, and can be separated from the still 
 liquid toluene. Similarly pure sulphuric acid is pre- 
 pared from the aqueous product obtained by evapora- 
 tion in platinum vessels. Quite recently this method 
 of purification by refrigeration has been applied on a 
 large scale to the purification of organic bodies, such 
 as ether, chloroform, &c., by Pictet and Liebreich. 
 
 So far we have dealt with what may loosely be 
 called mechanical methods of resolving mixtures. These 
 can, of course, also be resolved by chemical means, but 
 in this case one or other of the constituents of the 
 mixture necessarily suffers a permanent change. Thus, 
 a zinc compound in solution can be freed from an 
 admixed copper compound by passing sulphuretted 
 hydrogen through the acidulated mixture. The copper 
 compound is thus changed into insoluble copper 
 sulphide, which can be filtered off, leaving the zinc 
 compound in solution. 
 
 A number of very subtle chemical methods of re- 
 solving mixtures of very closely allied bodies have 
 recently been introduced under the generic title " frac- 
 tionation." These methods will receive more detailed 
 notice when we speak of the elements. 
 
 It will now be clear that in a mixture the constituents 
 
CHEMICAL CLASSIFICATION 53 
 
 conserve the properties peculiar to themselves in the 
 free state. Hence it follows that a mixture when 
 subjected to certain operations shows in general dif- 
 ferential results part dissolves, part does not; part 
 sinks, part floats so that it can be sorted into its 
 constituents by taking advantage of sufficiently pro- 
 nounced differences in degree of a common property of 
 its constituents. In thus studying the resolution of 
 mixtures, we have gained some knowledge of the more 
 general and important methods employed by chemists 
 for purifying substances. 
 
 Turning now to homogeneous substances. These, as 
 was first hinted by Boyle, can be divided into the two 
 classes, elements and compounds. The elements are 
 practically the units of chemistry ; according to definite 
 laws, they unite in manifold combinations and permuta- 
 tions to form the now almost infinite number of known 
 compounds ; but they defy all attempts to resolve them 
 into dissimilar constituents. Mercury, e.g., cannot in 
 any known way be split up into dissimilar portions, 
 each portion weighing less than the mass started with. 
 In short, elements constitute that class of homogeneous 
 substances which is amenable to synthesis only; ele- 
 ments cannot be analysed. They constitute what may 
 metaphorically be called the alphabet of the science. 
 
 But if compounds are composed of and are resolvable 
 into elements, how, it may be asked, do we distinguish 
 them from mere mixtures? Put briefly, the distinction 
 is this : mixtures can be resolved by processes of mere 
 sorting; compounds cannot be so resolved. When 
 
54 LAW AND THEORY IN CHEMISTRY 
 
 elements combine chemically, the properties of the 
 resulting compound are only remotely functions of the 
 properties of the elements combining. The altogether 
 new properties of the compound are not by any means 
 the combined properties of the constituents ; they show 
 no partial or differential character. A compound is 
 either wholly soluble or wholly insoluble in a given 
 menstruum ; it is either wholly magnetic or wholly non- 
 magnetic ; under the highest magnifying power attain- 
 able it is optically homogeneous. A knowledge of the 
 salient physical properties of the free elements consti- 
 tuting a compound does not assist us in our endeavours 
 to resolve that compound into its constituents, as is the 
 case with mixtures. A mixture can be resolved into 
 its constituents without involving a simultaneous and 
 permanent change in the matter separated or other 
 matter ; whereas the analysis of a compound in general 
 demands the permanent change of at least one con- 
 stituent of the compound as well as of extrinsic 
 matter. Thus with respect to the preparation of 
 copper from its compound copper oxide, the knowledge 
 that copper is a comparatively non-volatile insoluble 
 body, while oxygen is a soluble gas, avails us nothing ; 
 and the realisation of the project absolutely demands 
 the simultaneous and permanent change of some 
 such substance as carbon or hydrogen into oxidised 
 products. 
 
 Though the distinction between mixture and com- 
 pound is in theory easily enough grasped, yet it cannot 
 be denied that in practice it is often extremely difficult, 
 
CHEMICAL CLASSIFICATION 55 
 
 if not impossible, to state definitely whether in particular 
 cases we are dealing with the one or the other class 
 of matter. Physiological chemistry furnishes us with 
 several examples of this difficulty. Is haemoglobin a 
 " chemical unit " (i.e., homogeneous), or is it a mere 
 mixture of the substances globin and hsematin, which 
 it so readily yields? Is egg albumen the most 
 thoroughly investigated of the proteids a true 
 chemical compound, or is it a mixture of several 
 individual but closely allied proteid substances? Are 
 the varying results of its analyses made by different 
 experimenters to be ascribed solely to accompanying 
 difficultly removable impurities, or to the fact that it is 
 essentially a variable mixture ? Is gluten the sticky 
 constituent of dough a definite substance, or is it a 
 mixture of the proteids pre-existent in flour but 
 changed by ferment action ? To such questions physi- 
 ologists cannot at present vouchsafe definite answers; 
 indeed it may be said that one of the most formid- 
 able obstacles to the rapid advance of physiological 
 chemistry is the uncertainty as to whether many of 
 the fundamental substances with which it deals, and 
 which have received definite names, are, as a matter 
 of fact, " chemical units " (i.e., chemically pure 
 bodies). 
 
 Again, the question is often asked, is a solution of. 
 salt in water really a mere mixture of salt and water 
 particles, or does it contain what may be called low- 
 grade unstable compounds of salt particles and water 
 particles disseminated throughout the general bulk of 
 
56 LAW AND THEORY IN CHEMISTRY 
 
 the solution ? Is solution a purely physical change 
 in aggregation, or does it involve true chemical com- 
 binations between the particles of the solvent and 
 those of the dissolved substance ? 
 
 The investigation of the nature of solution takes 
 a most prominent place in the chemistry of to-day. 
 While some chemists are attacking the subject from 
 one side by purely physical methods, and summing 
 up their results in terms of partial and provisional 
 theories involving only physical ideas, others are assail- 
 ing the subject from a purely chemical standpoint. 
 Yet this statement must not be interpreted as assert- 
 ing the existence of two rival and mutually exclusive 
 schools on the question of solution. All chemists are 
 pretty well agreed that solution is neither a simple 
 physical nor a simple chemical change, but that it 
 is a very complex process involving simultaneous 
 physical and chemical changes. The exact point at 
 issue is the actual state of dissolved substances, i.e., 
 the constitution of solutions. 
 
 The only satisfactory explanation that has as yet 
 been given of the results of investigations of solutions 
 from their physical aspect, makes the solvent a passive 
 medium in which the discrete molecules of the dis- 
 solved substance (or the products of dissociation of 
 these molecules, the ions) uniformly distribute them- 
 selves. From this point of view the solvent bears 
 much the same relation to the dissolved substance 
 as the space defined by the walls of a vessel does 
 to the gas it contains ; solution, in fact, shows an 
 
CHEMICAL CLASSIFICATION 57 
 
 analogy to the evaporation of a liquid in a confined 
 space. 
 
 On the other hand, investigations carried out from 
 a chemical standpoint seem to demand the recog- 
 nition of a definite chemical interaction between, as 
 distinguished from a mere interpenetration of, the 
 molecules of the dissolved substance and those of the 
 solvent. When a salt is dissolved in water it appears 
 as if the salt molecules combine with the water mole- 
 cules to form definite yet very unstable chemical com- 
 pounds true liquid hydrates. 1 
 
 So far then it is clear that the two lines of investiga- 
 tion lead to different views regarding the structure of 
 solutions and the condition of the dissolved substance. 
 But it should be remembered that the more or less 
 mechanical conception of solution has arisen from the 
 exclusive study of very dilute solutions, while the 
 investigations leading to a more chemical rationale 
 have been chiefly carried out with strong solutions. 
 
 But little doubt can be entertained that future re- 
 search in this field will either establish a continuity 
 and relationship between these confessedly partial and 
 apparently diverse hypotheses, or by so modifying 
 one or both, will fuse them into a consistent and 
 comprehensive theory of one of the most common 
 but least understood of physico-chemical phenomena 
 solution. 
 
 Twilight is distinct from daylight, but one would 
 
 1 The constitution of solutions according to this view will be more 
 fully developed in the chapter on chemical equilibrium. 
 
 UNIVERSITY 
 
 oi_ 
 
58 LAW AND THEORY IN CHEMISTRY 
 
 hardly undertake to say where the one begins and the 
 other ends. So is it with the distinction between mix- 
 tures and compounds. The extremes of each class 
 contrast strongly enough; but through the inter- 
 mediacy of what are called molecular compounds 1 
 a convenient term with as yet a very vague connota- 
 tion the one class merges imperceptibly into, and 
 blends with, the other without break of continuity. 2 
 
 The following rough analogy may serve to fix in the 
 mind the fundamental differences between mixtures on 
 the one hand, and the two classes of homogeneous 
 bodies on the other. 
 
 Imagine four heaps, the first of needles only, the 
 second of pieces of thread only, the third of needles 
 and threads, and the fourth of threaded needles the 
 threads being knotted so as to form loops. The first two 
 heaps symbolise elements, say oxygen and hydrogen. 
 The third heap symbolises a mixture, say electrolytic 
 gas. From this heap unlike heaps can be made by 
 mere sorting. The fourth heap resembles a compound, 
 say, water, for while mere sorting alone will not resolve 
 this heap into two dissimilar ones, yet an isolation of 
 its constituents will be practicable after the destruction 
 by some extrinsic agency of the continuity of the thread 
 loops in the one case, or of the eyes of the needles 
 in the other. 
 
 1 See pages 144, 155. 
 
 2 In very interesting communications to the Chemical Society's 
 Journal, vol. Ixi. p. 114, Picton and Linder have shown that there is 
 a continuous series of grades of solution passing without break from 
 suspension to crystallisable solution. 
 
CHEMICAL CLASSIFICATION 59 
 
 But to return to the elements which have also con- 
 stituted a field for much of the more recent chemical 
 research. The list of elements has undergone many 
 vicissitudes since its first establishment. Until the 
 year 1811, when Davy introduced the powerful ana- 
 lytic method of electrolysis into chemistry, the alkalis 
 potash and soda, and the alkaline earths, baryta, 
 strontia, and lime, were regarded as elements. On 
 the other hand, chlorine up till this date was regarded 
 as a compound, being the oxide of a hypothetical 
 element. The establishment of the compound nature 
 of the alkalis and alkaline earths, and the simple nature 
 of chlorine was attended by important modifications 
 in the reigning theories of chemistry. Nitrogen was 
 for a time deprived of elementary dignity, being 
 regarded as the oxide of an unknown element nitri- 
 cum, and even the elementary nature of such un- 
 doubtedly simple substances as hydrogen, phosphorus, 
 and sulphur, was for a time called into question by 
 Berzelius. 
 
 It is perfectly impossible to state definitely the num- 
 ber of the elements ; indeed, the practical proof of the 
 simplicity of a given matter is one of the hardest tasks 
 that falls to the lot of the chemist. Lavoisier's list 
 only contained twenty-six members, including what 
 are now known to be forms of energy, heat and 
 light, but most chemists are now agreed that the 
 number of unequivocal elements lies somewhere be- 
 tween sixty-six and seventy. This uncertainty arises 
 from the fact that in recent years it has been found 
 
60 LAW AND THEORY IN CHEMISTRY 
 
 necessary to open a suspense list for the temporary 
 accommodation while sub judice of the many new 
 claimants for elementary distinction, and opinion is 
 divided as to the exact number of elements which 
 ought to be suspended. Even when two naturalists 
 agree as to the absolute number of unequivocal elements, 
 it does not follow that they favour exactly the same 
 elements throughout. One may favour the advance- 
 ment of an element A from the suspense to the estab- 
 lished list, while the other may think the evidence 
 incomplete in the case of A, but satisfactory in the 
 case of another claimant B. 
 
 Many of the elements constituting this suspense or 
 probationary list are not known in the free state, but 
 only in combination with oxygen in the form of very 
 refractory \i.e., difficultly reducible) oxides. Eefractory 
 oxides are generically known as earths, and as the 
 particular oxides under consideration are, though 
 widely diffused, only met with in nature in very small 
 quantities, they are usually referred to collectively as 
 the rare earths. Many of these oxides, and therefore 
 presumably the metals which they contain, are so 
 similar in their chemical properties that they can 
 neither be separated nor distinguished from one another 
 by ordinary chemical operations or tests. It is not 
 improbable that some of them will turn out on nicer 
 investigation to be either variable mixtures or allotropic 
 forms, while others may prove themselves compounds 
 of already known elements and so share the fate of 
 Bergman's " siderum," which turned out to be nothing 
 
CHEMICAL CLASSIFICATION 61 
 
 else than iron phosphide, and Eichter's " nickolanum," 
 which was merely an impure nickel. 
 
 The most important sources of these rare and doubt- 
 ful elements are the minerals gadolinite, cerite, and 
 samarskite. 
 
 From gadolinite, Bunsen and Bahr isolated the two 
 new earths yttria and erbia. These they regarded as 
 the pure oxides of new elements which they called 
 yttrium and erbium, but they did not isolate these 
 elements. Their yttria was pale yellow in colour. 
 
 Soon after this isolation, Smith and Delafontaine 
 separated a much darker yellow yttria from samarskite. 
 They therefore concluded that it was impure, and suc- 
 ceeded in separating from it a new orange-coloured 
 earth, terbia. This left a residue of perfectly white 
 yttria. Hence they concluded that Bunsen and Bahr's 
 yttria was non-homogeneous. Nor did Bunsen and 
 Bahr's erbia prove itself a simple substance, for 
 Marignac succeeded in resolving it into what he called 
 true erbia and ytterbia. Mlson then found that 
 ytterbia was non-homogeneous, and split it up into 
 true ytterbia and the earth scandia predicted by 
 Mendeleeff. Cleve, in making a spectroscopic study 
 of Marignac's erbia, came to the conclusion that it 
 was not homogeneous, but was a mixture of true erbia 
 and two new oxides, holmia and thulia. Finally, de 
 Boisbaudran brought forward reasons for suspecting 
 the integrity of holmia, which he ultimately resolved 
 into true holmia and dysprosia. This little sketch 
 does not include all nor nearly all the rare earths; 
 
62 
 
 LAW AND THEORY IN CHEMISTRY 
 
 it is intended to be illustrative rather than compre- 
 hensive and encyclopaedic. 
 
 Bunsen and Bahr Yttria. Erbia 
 
 Smith and Delaf ontaine . 
 Marignac .... 
 Nilson .... 
 
 Clfcve 
 
 Boisbaudran 
 
 Yttria. Terbia. 
 
 Erbia. 
 
 Ytterbia. 
 
 I 
 
 I | 
 
 Ytterbia. Scandia. 
 
 Erbia. Holmia. Thulia. 
 
 Holmia. Dysprosia. 
 
 Thus Bunsen and Bahr's original alleged simple earths 
 yttria and erbia were far from being homogeneous, as 
 they believed ; they were, according to present views, 
 mixtures of at least seven discrete oxides. 
 
 The methods which have been chiefly used in re- 
 solving these earths are fractional precipitation of 
 solutions of the oxides in acids by ammonia ; fractional 
 decomposition by heat of the mixed nitrates ; and 
 fractional crystallisation as already exemplified in the 
 case of potassium chlorate and chloride. 
 
 If to a solution which contains two earths of different 
 basicity (and all the earths differ to a greater or less 
 degree in this property) insufficient ammonia be added 
 to precipitate them both, then the less basic earth yields 
 more completely to the precipitant, and can be filtered 
 off, while the more basic constituent resists its action 
 and remains in solution. Thus after several thousands 
 of such fractional precipitations conducted in an orderly 
 
CHEMICAL CLASSIFICATION 63 
 
 manner, the two earths may be completely separated 
 from one another so far as ammonia can effect this 
 separation. The separated products are then ready for, 
 and should be subjected to, some other kind of frac- 
 tional resolution. If they yield to no other fractional 
 methods, then they are pro tern, elements. 
 
 As regards separation by fractional decomposition of 
 nitrates, it should be mentioned that here again the 
 nitrate formed from the less basic earth yields more 
 easily to the decomposing action of heat than the 
 nitrate formed from the more basic earth. Let A and 
 B be two earths, of which A is more basic than B. 
 Form the nitrates of these earths by acting on them 
 with nitric acid, and then partially decompose by heat 
 the mixed nitrates into their corresponding oxides. 
 Take up the residue with water and filter. The residue 
 will consist chiefly of the oxide of B and the filtrate of 
 the undecomposed nitrate of A. 
 
 In addition to the rare earths already mentioned, the 
 following, among others of very doubtful claims, are 
 candidates for the class of simple substances : yttrium, 
 ytterbium, zirconium, thorium, decipium, lanthanum, 
 samarium, cerium, praseodymium, neodymium. 
 
 The oxides of all the substances so far mentioned 
 have been isolated, and the elements of which they 
 are oxides have had provisional numbers assigned for 
 their atomic weights. These numbers are necessarily 
 affected with more or less uncertainty, for in many 
 cases they depend on arbitrarily assumed formulae for 
 the oxides. 
 
64 LAW AND THEORY IN CHEMISTRY 
 
 But even the admission of all the foregoing sub- 
 stances into the " elementary hierarchy " does not satisfy 
 some chemists, who, either from the spectroscopic ex- 
 amination of minerals from different sources, or from 
 the results of long-continued and varied fr action ations 
 in the laboratory, have come to the conclusion that 
 most of the above alleged elements are still mix- 
 tures. These chemists have not, however, succeeded 
 in isolating the constituents of these mixtures ; 
 they only find, according to their interpretation of 
 spectroscopic phenomena, indications of non-homo- 
 geneity. 
 
 Thus Crookes believes yttrium to be a mixture of 
 at least five elements, and Kriiss and Nilson would 
 find at least nine elements in the once reputed element 
 didymium, not two, praseodymium and neodymium, as 
 Welsbach thinks. But it should be here emphasised 
 that even could these elements be isolated they would 
 not be distinguishable from one another chemically. 
 Their discrimination would demand the subtile search- 
 ing power of the spectroscope, and our most refined 
 experimental methods would fail to establish conclusively 
 any difference in their atomic weights. As yet, these 
 " meta-elements," as they have been called by Crookes, 
 are mere possibilities, not certainties, and it may be 
 said that their substantiation, while of great specula- 
 tive and theoretic interest, would not at all influence 
 ordinary practical chemistry, which would still continue 
 to regard yttrium, for instance, not its five meta- 
 elements, as its working unit. It should also be 
 
CHEMICAL CLASSIFICATION 65 
 
 remembered in this connection that as yet spectroscopic 
 science "is still, for want of laws, at the epoch of 
 accumulation of facts, not of their possession." It 
 is as if chemists were making their preconceived 
 notions of the nature and number of the elements to 
 fit in with spectroscopic phenomena, instead of first 
 thoroughly establishing principles of spectroscopy, and 
 then using these to interpret the results of their investi- 
 gations on the elements. 
 
 Generally speaking, the upholders of the Periodic 
 Law l perhaps the greatest chemical generalisation of 
 modern times are averse to the idea of any great 
 increase in our commonly accepted number of elements. 
 In the tabular scheme which Mendele'eff gives as the 
 expression of this law, only six of the rare metals 
 are admitted, yttrium, ytterbium, lanthanum, cerium, 
 thorium, and zirconium, and there are only, so to 
 speak, vacancies left for a very limited number of other 
 elements, unless indeed the new elements have atomic 
 weights greater than that of uranium, which has the 
 highest atomic weight (240) of all known elements. 
 None of the above claimants would fall into this 
 category, and it seems very improbable that many, if 
 any, elements with atomic weights greater than 240 
 remain to be discovered. 
 
 This recent work on the rare earths, combined with 
 the results of the application of the spectroscope to 
 the investigation of the stars and nebulae, has resulted 
 
 1 See article, " Periodic Law," in Muir and Morley's new edition of 
 Watts' Dictionary of Chemistry. 
 
66 LAW AND THEORY IN CHEMISTRY 
 
 in modern views regarding the nature of the elements 
 which approximate to those held by the Aristotelian 
 school, in so far as they postulate one fundamental 
 matter. No longer are the elements universally be- 
 lieved to be ultimate and independent in their nature, 
 as Boyle and Lavoisier insisted, but the belief is 
 growing that they consist of one fundamental matter 
 in various stages of condensation, and that the stage 
 of condensation is dependent on the temperature, being 
 less the higher the temperature a belief which re- 
 invests the alchemistic attempts at transmutation with 
 a certain amount of warranty. In 1815 Prout suggested 
 that this fundamental matter was hydrogen, but the 
 suggestion was shown to be incorrect by the classical 
 work of Stas and others. At present the funda- 
 mental matter called "protyle," without any further 
 attempt being made to characterise it, is, so to 
 speak, placed lower down in the scale than hydrogen, 
 which is itself regarded as protyle in a certain ad- 
 vanced stage of condensation. 
 
 Crookes regards all elements as having been gradu- 
 ally evolved from this protyle; the heavier elements, 
 such as bismuth, thorium, and uranium, being the 
 younger species ; the lighter elements, such as lithium 
 boron, &c., being the older. A consequence of the 
 adaptation of this theory to the teaching of the 
 periodic law, is that what we are accustomed to call 
 elements ought really to be called " elementary 
 groups " ; that what we are accustomed to regard as 
 the atomic weight of an element does not accurately 
 
CHEMICAL CLASSIFICATION 67 
 
 represent the weight of each atom constituting that 
 element, but the mean of the weights of the atoms 
 making up a definite mass of the element, these 
 atoms being of slightly different weights. Thus the 
 atoms constituting a mass of yttrium have not all 
 identically the same weight ; while some, it may be 
 the greater quantity, weigh 89 times as much as the 
 average weight of the hydrogen atom, a few weigh 
 89-01, a few others again 88*99 times as much, and 
 so on. As before stated, Crookes believes that there 
 exist in the " element " yttrium, at least five different 
 kinds of atoms, each constituting what he calls a 
 meta-element. These meta-elements are regarded as 
 protyle in five very slightly different stages of con- 
 densation. He is not quite decided as to whether 
 one of these meta-elements largely preponderates over 
 the other four as above hinted, but he rather inclines 
 to this yiew. If such is the nature of the yttrium 
 commonly accepted as an 'element, what, he asks, is 
 there improbable in the idea that minute investigation 
 would result in the assigning of a similar structure to 
 the more common and familiar elements to calcium 
 and barium, for instance ? 
 
 If we accept this bold application of the principle 
 of evolution to the elements, we are scarcely justified 
 in any longer speaking of matter as inanimate. Accord- 
 ing to Crookes, the atoms of the elements are born, 
 undergo secular vicissitudes, and die; and the rare 
 elements are, in his opinion, strictly comparable to rare 
 plants and animals, in that they are simply examples 
 
68 LAW AND THEORY IN CHEMISTRY 
 
 of special kinds of atoms that have failed to harmonise 
 with their environment. 1 
 
 1 For further details on this abstruse and highly speculative subject, 
 reference must be made to Crookes' original papers in the Journal of 
 the London Chemical Society, vols. liii. and Iv. In this connection 
 Clerk Maxwell's article " Atom " in the Encyclopaedia Britannica should 
 also be read. In this article reasons are advanced for the diametri- 
 cally opposed belief that the atoms of any one element, hydrogen, for 
 instance, are eternal and identically similar among themselves, bearing 
 the impression, so to speak, of articles stamped by machinery. 
 
CHAPTER IV. 
 
 THE ATOMIC THEORY. 
 
 FROM the beginnings of philosophy the question of the 
 constitution of matter 1 has exercised the minds of 
 philosophers. Defining matter as that which occupies 
 space, it has been asked, does matter completely fill 
 the space its outline occupies or not ? Does the gas 
 contained in a flask completely fill the space it occupies, 
 i.e., the space defined by the sides of the flask, in the 
 same way as to our gross senses jelly fills a cup ? Or 
 does it, on the contrary, fill the space in much the 
 same way as apples fill a barrel, so that with sufficiently 
 magnified vision we could plunge an exceedingly delicate 
 style into the flask in such way that the style's point, 
 occupying at a given instant an interstice between 
 the individual particles of the gas, would not at that 
 instant actually touch the gas ? 2 
 
 1 The implications of the terms constitution of matter and nature of 
 matter must be carefully distinguished. The ultimate nature of matter 
 is a metaphysical rather than a chemico physical question. For a 
 resume of the varied views held on this subject see Tait, Properties of 
 Matter. 
 
 2 It should be remarked that the question of the structure of matter 
 is here being developed from the conventional chemical standpoint, 
 not from the phy.sical. The conception which most chemists, I think, 
 have of a piece of chalk, for instance, is a collocation of molecules, each 
 made up of hard atoms, each having the incompletely defined composi- 
 
 69 
 
70 LAW AND THEORY IN CHEMISTRY 
 
 Stating the question otherwise ; is apparently con- 
 tinuous matter really continuous and capable of in- 
 finitely small division without losing its individuality, 
 or is it really discontinuous and capable of division 
 down only to the limit marking the beginnings of dis- 
 continuity ? Are geometrical conceptions or arith-* 
 metical to underlie our theories of the constitution of 
 matter ? 
 
 The early Greek philosopher Leukippos (428 B.C.) 
 was the first to make answer in a vague way to these 
 questions. 1 He decided that matter is not continuous, 
 
 tion x CaC0 3 , and each separated from its neighbours by " void." But 
 physicists, whilst admitting the ultimate heterogeneousness or grained- 
 ness of apparently homogeneous matter, do not deny its perfect con- 
 tinuity. The heterogeneity may take the form of alternations of matter 
 and void, but it may involve only such differences as exist; between 
 solid and fluid, or between substances differing enormously in density ; 
 or such heterogeneousness as differences in velocity and direction of 
 motion, in different positions of a vortex ring in a homogeneous liquid. 
 Three points should be noted in this connection (1) in discussing the 
 constitution of matter physicists employ the terms atom and molecule in 
 a very loose way ; (2) their connotation of the term heterogeneity is 
 quite different from that which the same term bears in Chap. III. ; and 
 (3) although in the restricted domain of chemistry we make provisional 
 use of the conception and term atom, yet from a broadly philosophic 
 standpoint we do not assert that the chemical atom represents the 
 absolute limit to the divisibility of matter. The sociologist adopts as 
 his unit the individual, quite regardless of the anatomist's powers. 
 The atom is the chemical individual. See Lord Kelvin, Popular 
 Lectures and Addresses, vol. i., and Maxwell's article "Atom " in the 
 Encyclopcedia Britannica. 
 
 1 Later antiquity chiefly studied the atomic theory in the more 
 elaborated form it attained under Demokritos, the pupil of Leukippos, 
 and thus the disciple has pretty well usurped the place of the master in 
 history. For a very concise account of the passage of Eleaticism into 
 Pluralism, thence into Atomism, with the fundamental changes of idea 
 this passage involved, see Burnet's Early Greek Philosophy. 
 
THE ATOMIC THEORY 71 
 
 but is made up of numerous small separate and in- 
 divisible particles called atoms. His views, embraced 
 and developed by the Epicurean school, have been pre- 
 served for us by the Roman poet Lucretius (B.C. 99) in 
 his "De rerum natura." Here is the main argument, 
 as presented in the poem. First comes the perfectly 
 gratuitous assumption that reproduction the agglo- 
 meration of scattered, invisible particles into visible 
 bodies is slower than decay, the breaking up of bodies 
 into invisible particles. From this premise it follows 
 that there must be a limit to the breakage ; else the 
 decay of infinite past ages would have left nothing 
 visible and tangible. This limit to breakage or decay 
 involves the idea of a least in things, which least is 
 called an atom. Lucretius also taught that there were 
 atoms of many different forms and weights an infinite 
 number of each kind ; and that all change consisted in 
 the intimate contact or separation from such contact of 
 these different kinds of atoms. This view of matter as 
 presented by Lucretius is more of a vague poetic cosmo- 
 gony than a scientific theory. It had no numerical 
 conciseness about it, and was altogether too indefinite 
 to be suggestive. It bears to the atomic theory, as 
 accepted at present, about the same relation as do the 
 scientific dreams and guesses of a Goethe to the experi- 
 mentally substantiated inspirations of a Newton. 
 
 Passing over the use which Newton made of an 
 atomic conception of the atmosphere to make his 
 theoretically derived velocity of sound tally with the 
 experimental value a physical rather than a chemical 
 
72 LAW AND THEORY IN CHEMISTRY 
 
 point we do not find any serious attempt at a strictly 
 scientific theory of the constitution of matter till we 
 come to the year 1804. In this memorable year 
 John Dalton first propounded his atomic theory a 
 theory which is so fundamental in the chemistry of 
 to-day, that were it in any way disproved, the whole 
 superstructure of the science would fall in hopeless 
 and chaotic ruin. 
 
 Lord Kelvin has said that only when we can measure 
 the thing we are speaking about and express it in 
 numbers do we begin to know really anything about 
 it. Judged by this sentiment, Dalton brought us to 
 the beginnings of a true knowledge of atoms, for as 
 we shall see his theory was nothing else than that of 
 the early Greek philosophers mentioned, founded not 
 on fancy, but on the careful observation of the facts 
 of nature, and reduced to numerical definiteness. Let 
 us see how Dalton was led to embrace the atomistic 
 or limited divisibility view of matter. 
 
 The fact that water does not dissolve all gases in 
 the same quantities seems first to have suggested to 
 Dalton the view that gases consist of large numbers 
 of discrete particles, but what in his opinion chiefly 
 demanded a resuscitation of the atomic theory was 
 undoubtedly his discovery of the law of multiple pro- 
 portions taken in conjunction with the previously sub- 
 stantiated law of definite proportions. Hence we must 
 first glance at these two fundamental chemical laws. 
 
 The discovery of the law of definite proportions 
 cannot be ascribed to any particular individual ; it 
 
THE ATOMIC THEORY 73 
 
 was arrived at simultaneously by several chemists soon 
 after Lavoisier introduced quantitative methods into 
 the science. If A, B, &c., represent the masses in 
 which definite homogeneous substances interact to 
 produce new homogeneous substances represented 
 quantitatively by C, D, &c., then this law in its most 
 general form states that no matter what the con- 
 ditions of reaction or the absolute values of A, B, 
 
 A A A B B C 
 C, D, &c., may be, the ratios -, - -, -, -, , &c., 
 
 > U JJ L> _U 1) 
 
 are all constant. A particular case is when A and B, 
 &c., represent elements uniting to form the single 
 compound C, and the law is generally enunciated in 
 terms of this special case as follows : no matter the 
 conditions of formation, or the proportions in which 
 the constituents are taken, a given chemical compound 
 has always exactly the same percentage composition. 
 
 Most laws (e.g., the gaseous laws of Boyle and Charles) 
 are only true under specified conditions or between 
 certain narrow limits, but the law now under observa- 
 tion is as far as we can tell an absolute and uncon- 
 ditional one. Working with the haloid salts of silver 
 preparing them in many different ways and from 
 different proportions of the ingredients Stas proved 
 the law accurate to 1 part in 10,000 parts, 1 an accuracy 
 
 1 An ordinary chemical balance is capable of an accuracy of 1 in 10 6 , 
 and Miller in his elaborate construction of the standard pound attained 
 an accuracy of 1 in 3 x 10 8 , but in Stas' experiments errors considerable 
 in magnitude when compared with those involved in mere weighing, are 
 incidental to the operations of filtering, washing, incinerating, &c. An 
 accuracy of 1 in 10,000 is about that attainable in a simple measure- 
 ment of length. 
 
74 LAW AND THEORY IN CHEMISTRY 
 
 of high order when we recall the diverse operations 
 involved. 
 
 Yet the law of definite proportions has had its vicissi- 
 tudes. In the beginning of this century Berthollet 
 denied its truth as above enunciated, asserting that 
 the composition of a body might vary within certain 
 limits, this variation being dependent on the masses in 
 which its constituents were allowed to react l just as 
 recipes with slight quantitative differences will give to 
 all intents and purposes the same culinary product. 
 In this mistaken view he was successfully opposed by 
 Proust, who was led to a warm espousal of the law 
 by the following experiment. Proust analysed native 
 malachite and found that it gave 71 "9 per cent, of cupric 
 oxide. He then dissolved the mineral in acid and repre- 
 cipitated the malachite in an amorphous form. This 
 artificial product gave on analysis 71 '9 per cent, of 
 cupric oxide. Hence Proust concluded that no matter 
 whether malachite is formed in the crystalline condition 
 in the earth's crust, where it is subjected to the influ- 
 ence of mighty secular processes of which we know next 
 to nothing, or whether it is formed artificially and sud- 
 denly in the laboratory, it has in each case identically 
 the same composition. He then showed that the 
 alleged variable nitrate of mercury which Berthollet 
 brought forward to vindicate his attitude, was not a 
 
 1 The influence of mass in causing variations in the relative quantities 
 of the products of a change (not of their individual compositions as 
 Berthollet supposed) lies at the foundation of the study of chemical 
 affinity in the modern acceptation of this term. See Pattison Muir, 
 Principles of Chemistry. 
 
THE ATOMIC THEORY 75 
 
 homogeneous body but a mixture of mercurous and 
 mercuric nitrates, in proportions varying according to 
 the relative masses of mercury and nitric acid brought 
 together. Thus arose the distinction between homo- 
 geneous substances and non-homogeneous substances 
 or mixtures detailed in Chap. III. 
 
 Passing on now to the law of multiple proportions. 
 Before Dalton's time the fact had been recognised that 
 a given element A will combine with another element 
 B in two or more proportions to form two or more 
 compounds, for each of which the law of definite pro- 
 portions is true. But up to the year 1803, owing to 
 the customary method of stating the results of analysis, 
 no regularity had been detected in the values of the 
 combination masses of A and B as between compound 
 and compound. Dalton analysed the two compounds 
 of carbon and hydrogen, methane and ethylene; as 
 also the two compounds of carbon and oxygen, carbonic 
 oxide and carbonic anhydride, with these results 
 
 ETHYLENE. 1 METHANE. 
 
 = 85-71 p.c. H = 14-28p.c. = 75 p.c. H = 25 p.c. 
 
 or approximately or approximately 
 
 6 C! 6 
 
 CARBONIC OXIDE. CARBONIC ANHYDRIDE. 
 
 = 42-86 p.c. O = 57'24p.c. C = 27'27p.c. = 72'72p.c. 
 
 or approximately or approximately 
 
 _ 6 ? = J> 
 
 = 8 16 
 
 1 The percentages here given are not the values actually obtained 
 by Dalton. 
 
76 LAW AND THEOEY IN CHEMISTRY 
 
 No regularity appears between the valves for the two 
 hydrides on the one hand, or the two oxides on the 
 other, until we consider a constant mass of either of 
 the common elements in the two cases, instead of ex- 
 pressing the results in the conventional percentage 
 form. It is then noticed that the amount of hydrogen 
 combined with six unit masses of carbon in marsh 
 gas is exactly 1 double (not 2 J times, nor 1 T Vy times, 
 but exactly double) the amount of hydrogen combined 
 with six unit masses of carbon in ethylene. Similar 
 observations apply to the oxides of carbon, and indeed 
 to all cases where two elements combine in different 
 proportions to form different compounds. In every 
 case the following regularity is observed. If an element 
 A combines with an element B in two or more pro- 
 portions to form two or more compounds, and if in 
 these compounds we consider a fixed mass of either of 
 the elements, say A, then the masses of B combined in 
 the several compounds with this fixed mass of A, are 
 in general so related that the compounds richer in B 
 contain of B masses expressible as whole multiples 
 of the mass of B, contained in the compound 
 poorest in B. This is Dalton's law of multiple pro- 
 portions. 2 
 
 1 This exactness is never of course attained in practice, on account 
 of unavoidable errors of observation, but it is more and more nearly 
 realised as we perfect our analytic methods. Hence the assumption is 
 warranted that the relations in question are in reality absolutely exact. 
 The mathematical conception of a limit is here implicated. 
 
 2 Some apology for this diffuse enunciation seems called for. In the 
 majority of chemical text-books, the word multiple is never once intro- 
 duced in the enunciation of the law of multiple proportions, its place 
 
THE ATOMIC THEORY 77 
 
 What explanation 'can be adduced of these laws? 1 
 Dalton answered, the atomic theory of matter. Admit, 
 he said, that simple matter of every kind consists of 
 little indivisible particles called atoms, that the atoms 
 of one and the same kind of matter have always exactly 
 the same weight and the same properties, but that the 
 weight and other properties of the atoms differ from 
 substance to substance ; that chemical combination 
 consists in the coming together into intimate contact 
 of definite numbers of simple atoms, to form definite 
 compound atoms which in turn are all exactly similar 
 for one and the same compound, bearing, so to speak, 
 the impress of goods stamped by machinery ; then, 
 admitting all this, the fundamental laws of chemistry 
 
 being taken by the vague, elastic, yet polarised term "simple ratio." 
 In the statement of the law given in the text the attempt has been 
 made to remedy these defects, even at the expense of brevity. 
 
 1 In addition to the two laws given in the text, many add a third 
 the law of reciprocal proportions, or the law of combining weights. 
 As a matter of fact, the perception of the numerical relations embraced 
 by this law did not antedate and pave the way for the atomic theory, 
 but this theory established, certain numerical results necessarily fol- 
 lowing from it were grouped together under one or other of the above 
 titles, thus : if AB and AC represent quantitatively two compounds, 
 then any compound of B and C will be represented quantitatively by 
 ?iBmC, where n and m are whole numbers, and A, B, C, the com- 
 bining weights of the elements. Interpreted in terms of the atomic 
 theory, this law may be regarded as stating that the atoms of a given 
 element have a fixed and unalterable mass in all the compounds that 
 element forms. Agl and AglOs might each have fixed composi- 
 tions without the ratio i" being identical in the two cases as theory 
 demands. Stas, in proving to an accuracy of 1 in 10 7 that the 
 ratio --= is the same in both compounds, has placed the atomic theory 
 on a very exceptional experimenta'l basis. 
 
78 LAW AND THEOEY IN CHEMISTEY 
 
 follow as necessary consequences. 1 It is quite easy to 
 see how the law of definite proportions follows ; let us 
 examine in detail the sequence of the multiple pro- 
 portion law. 
 
 Suppose matter to be non-atomic, to be infinitely 
 divisible and absolutely continuous. Let the symbol 
 
 represent a definite mass (expressed numerically 
 
 by 6) of carbon on this supposition; and the symbol 
 |U represent under the same conditions a mass of 
 hydrogen, numerically expressed by 1. Then ethylene 
 
 might be represented as 6 m . Now, analysis in- 
 
 forms us that methane contains relatively more hydrogen 
 than does ethylene. According to the above supposi- 
 tion, the mass of hydrogen combined with 6 unit masses 
 of carbon in methane might quite easily be 1-1-, 1^, llj, 
 or 2yf, &c. times the mass of hydrogen combined with 
 6 unit masses of carbon in ethylene ; for hydrogen is 
 jelly-like and structureless, and we can imagine a 
 
 1 Stallo divides the phenomena of chemical change into three classes. 
 The first "embraces the persistence of weight and the combination in 
 definite proportions ; the second, the changes of volume and the evolu- 
 tion or involution of energy ; and the third the emergence of a wholly 
 new complement of chemical properties." He asserts that the atomic 
 hypothesis is in no sense an explanation of phenomena of the second and 
 third classes, nor does it fully explain those of class i. in the sense of 
 generalising them and reducing many facts to one. It accounts for 
 them by simply iterating the observed fact in the form of an hypo- 
 thesis. He admits, however, " the merits of the atomic hypothesis as 
 a graphic or expository device as an aid to the representative faculty 
 in realising the phases of chemical or physical transformation." See 
 Concepts of Modern Physics, chap. vii. 
 
THE ATOMIC THEORY 79 
 
 quantity of it of any mass whatever. Given 
 
 as representing ethylene, the composition e L| would, 
 a priori, be quite as probable for methane as would 
 
 But now let us take the atomic theory, and for the 
 sake of present argument assume that ethylene con- 
 sists of 1 atom of carbon weighing 6 units, and 1 atom 
 of hydrogen weighing 1 unit. Then, since methane 
 contains more hydrogen relatively than ethylene, it 
 follows that per 1 atom of carbon methane must 
 
 contain 2, 3, 4 n atoms of hydrogen ; 
 
 for atoms are the indivisible units of chemistry. 
 Since all atoms of the same simple substance have 
 exactly the same weight, the law of multiple pro- 
 portions necessarily follows ; the amount of hydro- 
 gen per given amount of carbon must necessarily 
 be in methane a whole multiple of its value in 
 ethylene. 
 
 To sum up the whole matter ; the existence of a 
 multiple law according to the non-atomic theory of 
 matter is a possibility among an infinity of other 
 equally plausible possibilities ; the odds against its 
 existence are infinitely great. But of an atomic theory 
 of matter, a multiple law is an absolutely necessary 
 consequence. That Dalton clearly saw this consti- 
 tutes, perhaps, his greatest claim on the memory of 
 posterity. 
 
8o LAW AND THEORY IN CHEMISTRY 
 
 We must now follow Dalton in his attempt to find 
 the weights 1 of the atoms of elementary bodies. He 
 did not, of course, hope to determine the absolute 
 weights of the atoms a determination of which 
 even the advanced chemistry of to-day is incapable. 
 All he attempted to do was to find the relative weights 
 of the atoms. He made the hydrogen atom his standard, 
 and arbitrarily called its weight one ; all other atomic 
 weights were stated in terms of this standard. Thus 
 the statement the atomic weight of oxygen is 16 
 simply means that the atom of oxygen is sixteen times 
 heavier than the atom of hydrogen, whose absolute 
 weight expressed, say in a fraction of a milligram, is 
 unknown. 
 
 As an aid to the foundation of a system of atomic 
 weights, Dalton framed a series of empirical rules. 
 We cannot dissect out the train of reasoning which 
 culminated in these rules. They are merely the ex- 
 pression of Dalton's own preconceived notions mere 
 guesses, limited only by the condition that compound 
 atoms are of simple rather than of complex structure, 
 being as a rule made up of very small numbers of 
 elementary atoms. Rule 1. If only one compound of 
 two elements can be obtained, the compound must be 
 assumed a binary 2 one, unless there be good cause for 
 
 1 In some recent text-books, atomic mass replaces the term atomic 
 weight. In view of the strict proportionality which exists between 
 mass and weight, and in view of the hold the terms weighing and 
 weight have in the speech of everyday life, this innovation seems to 
 me unnecessary. 
 
 2 A binary compound, according to Dalton, is one whose "com- 
 
THE ATOMIC THEORY 81 
 
 some other conclusion. Rule 2. If two compounds of 
 two elements exist, then one is a binary, and the other 
 a ternary, compound. Rule 3. If three compounds of 
 two elements exist, then one is a binary, and the 
 other two are ternary, compounds. Rule 4. If four 
 compounds of two elements are known we should 
 expect two of them to be ternary, one to be binary, 
 and the fourth to be quaternary. Other rules follow, 
 dealing with the specific gravities of binary, ternary, 
 &c. compounds ; but for these, and indeed for Dalton's 
 work on the atomic theory as a whole, the reader is 
 referred to Ostwald's Klassiker der exakten Wissen- 
 schaften, No. 3, or to the Alembic Club Reprints. 1 
 
 To illustrate the application of these rules, we may 
 take the case of the atomic weight of oxygen. To 
 Dalton only, one compound of hydrogen and oxygen 
 was known, viz., water. Hence by Rule 1 the com- 
 pound atom of water consisted of one atom of hydrogen 
 in combination with one atom of oxygen. Now it is 
 found by actual experiments that about 8 grams of 
 oxygen combine with 1 gram of hydrogen to form 
 about 9 grams of water ; or, what is the same thing, 
 
 8 1 
 
 that - milligrams of oxygen combine with mg. 
 
 of hydrogen to form ^ mg. of water. Let us sup- 
 pound atom" contains only two simple atoms. In the "compound 
 atom " of a ternary body there are three atoms, and so on. 
 
 1 These brochures are more generally attainable than the Memoirs of 
 the Literary and Philosophical Society of Manchester, in which Dalton's 
 papers originally appeared. 
 
 F 
 
82 LAW AND THEORY IN CHEMISTRY 
 
 pose, simply for the sake of clearness, that a hydrogen 
 
 atom weighs ^ mg. 1 
 
 1 8 
 
 3 mg. H combines with 3 mg. 0. 
 
 or 1 atom H ,, ,, 
 
 but, by Kule i., 1 atom H ,,1 atom 0. 
 
 o 
 
 Therefore 1 atom weighs - mg. 
 
 So if we agree to call the atomic weight of hydrogen 
 one, the atomic weight of oxygen will equal eight. In 
 a perfectly similar manner Dalton arrived at the result ; 
 atomic weight of nitrogen = 4f. But 16 (=8 X 2) 
 and 14 (= 4f x 3) are at present regarded as the 
 approximately correct values for the atomic weights of 
 oxygen and nitrogen respectively. Why these num- 
 bers are adopted in preference to Dalton's will not be 
 apparent till we have treated of Avogadro's hypothesis. 
 When we attempt to fix by Dalton's rules the atomic 
 weight of such an element as carbon which combines 
 with hydrogen, oxygen, &c., in more than one pro- 
 portion, we find ourselves on the horns of a dilemma. 
 According to Rule 2, either methane or ethylene is a 
 binary body, but the rule does not indicate any means 
 of assuring ourselves which is the binary body. If 
 ethylene be assumed binary, methane will be ternary. 
 Ethylene will be CH, methane CH 2 , and the atomic 
 weight of carbon will be 6. If, however, methane be 
 
 1 This very rough approximation is derived from data furnished by 
 the kinetic theory of gases. 
 
THE ATOMIC THEOEY 83 
 
 assumed binary, ethylene will be ternary. Under this 
 assumption, the formula for methane will be CH, that 
 for ethylene C 2 H, and the atomic weight of carbon 
 will be 3. 
 
 Which of these two values for carbon are we to 
 select? There is nothing at all in Dalton's rules to 
 guide us in such cases as these. Failing inspired arch- 
 chemists, one is as justified from the premisses in 
 maintaining that the atomic weight of carbon = 3 as 
 another is in championing the value 6. 
 
 This ambiguity in Dalton's rules led to great diffe- 
 rences of opinion respecting the atomic weights of 
 several of the elements, and ultimately the confusion 
 became so great that many advocated the renunciation 
 of the atomic theory altogether. The light which it 
 indisputably shed on some points did not in the opinion 
 of many compensate for the fundamental uncertainties 
 with which it was hampered, and which it was power- 
 less to resolve. It is, I think, worthy of note that 
 the very cases which in the first instance suggested to 
 Dalton the law of multiple proportions and the atomic 
 theory were exactly those cases in which his elaborated 
 system was found wanting, and which led to its un- 
 popularity and temporary rejection. 
 
 This indefiniteness, which trammelled the young 
 atomic theory, was finally resolved by the knowledge 
 accruing from a careful study of the physico-chemical 
 properties of gases, to which we now turn. 
 
 In 1805, Gay-Lussac and Humboldt, investigating the 
 constancy of the amount of oxygen in the atmosphere, 
 
84 LAW AND THEORY IN CHEMISTRY 
 
 employed an analytic method first proposed by Volta. 
 This method, which has become classical, consists in 
 mixing the air with hydrogen in a eudiometer, ex- 
 ploding the mixture by an electric spark, and then 
 from the contraction which ensues calculating the 
 amount of oxygen present. To apply this method, an 
 accurate knowledge of the contraction which occurs 
 when liquid water is formed from hydrogen and oxygen 
 is obviously essential. After very careful experiments, 
 Gay-Lussac and Humboldt found that one volume of 
 oxygen combines with two volumes of hydrogen to form 
 a drop of water, whose volume in comparison with the 
 volumes of the gases exploded is in general negligible. 
 
 2 vols. H + 1 vol. = water with practically negligible vol. 
 
 Hence it follows that 3 volumes of the properly mixed 
 gases contract to zero volume on explosion. In other 
 words, suppose 2 cubic feet of hydrogen and 1 cubic 
 foot of oxygen (measured at atmospheric pressure), 
 placed in an air-tight vessel 1 and then exploded, the 
 pressure in the vessel would sink from 15 Ibs. per 
 square inch to nearly zero value. I say nearly, for the 
 water formed would vaporise and exert a small pres- 
 sure (or as it is currently but unfittingly called, a 
 tension) dependent as regards magnitude on the tempe- 
 
 1 This experiment was actually performed by Cavendish, who was 
 the first to determine the composition of water by volume (see Chap. 
 II. p. 53). Cavendish introduced the gases, hydrogen and oxygen, 
 mixed in the proper proportions into a vacuous globe, exploded the 
 mixture, and found that the vacuum was re-established, so that several 
 successive charges and explosions could be effected with one initial- 
 exhaustion of the globe. 
 
THE ATOMIC THEORY 85 
 
 rature. From all of this we conclude that the amount 
 of oxygen in a mixed gas is equal in volume to one- 
 third the contraction caused by sparking after an excess 
 of hydrogen has been added. 
 
 Supposing the water formed from 2 vols. hydrogen 
 and 1 vol. oxygen is changed into the gaseous state 
 by heating it above 100 C. so as to form super- 
 heated steam ; or supposing the whole experiment is 
 conducted throughout at a high temperature, so that 
 the water formed never condenses, what volume-relation 
 would the water gas or steam bear to the volumes of 
 hydrogen and oxygen which formed it ? In answer to 
 this inquiry, Gay-Lussac and Humboldt found that the 
 volume of steam formed is exactly equal to the volume 
 of the hydrogen (or what is the same thing, to twice the 
 volume of the oxygen) exploded. 
 
 Here Gay-Lussac and Humboldt met with a quan- 
 titative fact which astonished them by reason of the 
 extreme simplicity of the quantities involved. Two 
 cubic feet of hydrogen combine with exactly one cubic 
 foot of oxygen, not with -9 or 1^, but according to Gay- 
 Lussac and Humboldt with exactly 1 cubic foot, of 
 oxygen to form, not 2 T ^Q, but exactly 2 cubic feet of 
 steam at the same temperature. 1 
 
 1 Although we assert in the text the exactness of the ratio 2:1, yet 
 it seems well nigh impossible to prove this exactitude experimentally. 
 Scott and E. W. Morley, employing all the refinements of modern 
 methods, have repeated Gay-Lussac and Humboldt's experiment with 
 the following result : 
 
 2*00245 vols. of hydrogen combine with 1 vol. of oxygen (Scott). 
 
 2-0023 (Morley). 
 
 However, in all the theoretical deductions which follow, any slight 
 
86 LAW AND THEORY IN CHEMISTRY 
 
 The question suggests itself, does a similar simple 
 relationship in the reacting volumes manifest itself in 
 other cases of gaseous combination ? It does. 
 
 1 vol. of hydrogen combines with exactly 1 vol. of chlorine 
 to give 2 vols. of hydrochloric acid gas. 
 
 1 vol. of nitrogen combines with exactly 3 vols. of hydrogen 
 
 to give 2 vols. of ammonia gas. 
 
 2 vols. of nitrogen combines with exactly 1 vol. of oxygen 
 
 to give 2 vols. of laughing gas. 
 
 1 vol. of carbonic oxide combines with exactly 1 vol. of 
 chlorine to give 1 vol. of phosgene gas. 
 
 Several other instances might be adduced of Gay- 
 Lussac's Law of Volumes, which states that in homo- 
 geneous gaseous reactions (i.e., reactions in which all 
 the factors and products of the change are gaseous) all 
 the volumes involved are in such simple relationship 
 that the ratios can in every case be expressed in terms 
 of the first six digits. 1 
 
 Just as the law of multiple proportions called forth 
 an explanation, so it was not long before the why 
 and wherefore of Gay-Lussac's law was under discus- 
 sion. The outcome of this discussion was the conjee- 
 deviations from simplicity which may possibly exist in the ratios of the 
 combining volumes of gases will be disregarded, and the law of volumes 
 will be accepted as strictly true. 
 
 1 It should be noticed that when the factors of a homogeneous 
 gaseous reaction are elements, the volume of the resulting compound 
 is, with very few exceptions, always 2, provided the volume equation be 
 throughout reduced to its simplest terms. The formation of phosphine 
 from its elements furnishes an example of the exceptions referred to. 
 
 P 4 + 6H 2 = 4PH 3 
 
 1 vol. + 6 vols. = 4 vols. 
 
THE ATOMIC THEORY 87 
 
 ture 1 accepted by several chemists, but first definitely 
 enunciated by Berzelius, that equal volumes of gases 
 contain equal numbers of atoms, and hence that the 
 atomic weights of gases are proportional to their 
 specific gravities. 2 In view of the identical behaviour, 
 qualitatively and quantitatively, of different gases when 
 subjected to pressure and temperature changes, this 
 conjecture seemed very plausible. The laws of Boyle 
 and Charles seem absolutely to demand some such iden- 
 tity in mechanism of the various gases as is implied in 
 Berzelius' conjecture. But apart from this considera- 
 tion, it cannot be denied that the human mind has a 
 peculiar and inherent bias for the uncomplicated, and 
 it is probable that the extreme simplicity of the explana- 
 tion was one of the most potent factors in securing for 
 it a general acceptance. 
 
 Dalton and his school, however, absolutely refused to 
 accept Berzelius' interpretation of the law of volumes, 
 maintaining that the atomic weights could be deter- 
 mined only from " the ponderable relation of elements 
 
 1 Berzelius' interpretation of the Law of Volumes seems scarcely to 
 merit the title of an hypothesis, which " is any supposition we make in 
 order to endeavour to deduce from it conclusions in accordance with 
 facts which are known to be real." 
 
 2 Since the specific gravities of oxygen and nitrogen, referred to 
 hydrogen as standard, are respectively 16 and 14, it follows that these 
 values were adopted by Berzelius for the relative atomic weights of 
 these elements. It should be remarked that at first Berzelius regarded 
 the data derived from the law of volumes as belonging to what he 
 called " volume atoms " or "elementary volumes." These he conceived 
 of as something fundamentally distinct from Dalton's atoms. After a 
 time, however, he returned to the Daltonian conception of atoms and 
 applied to these the results following from his hypothesis. See Wurtz, 
 The Atomic Theory, pp. 43-48. 
 
88 LAW AND THEORY IN CHEMISTRY 
 
 in combination," i.e., from purely analytic data. As an 
 argument against this interpretation, those cases of 
 gaseous combination which occur without change of 
 total volume were adduced. 
 
 1 vol. hydrogen combines with 1 vol. chlorine, forming 
 2 vols. hydrochloric acid gas. 
 
 Assume that the volume of hydrogen considered is so 
 small that it contains only one atom ; then, accepting 
 Berzelius' conjecture, the equal volume of chlorine 
 with which it combines will contain only one atom, and 
 the double volume of hydrochloric acid gas formed 
 will contain two compound atoms. Now each of these 
 compound atoms must contain at least one hydrogen 
 atom and one chlorine atom. Therefore the two com- 
 pound atoms together must contain at least two atoms 
 of hydrogen and two atoms of chlorine. But we only 
 started with one atom of hydrogen and one atom of 
 chlorine. Therefore matter has been created an 
 absurd conclusion in view of all the proofs that exist 
 of the conservation of matter. 1 
 
 Berzelius was bound to admit the justness of this 
 reductio ad absurdum, and in order to meet it emphasised 
 the fact that his so-called hypothesis only extended to 
 the elementary gases, and not to compound gases and 
 
 1 A neat but indirect proof of the principle of the conservation of 
 matter is furnished by the constancy of the length of the year. This 
 depends on the masses of the earth and sun. If the masses of the 
 earth were continually changing by reason of the chemical changes 
 taking place, then some alteration in the length of the year would have 
 been produced within historic time. 
 
THE ATOMIC THEORY 89 
 
 vapours. In this form he continued to make a neces- 
 sarily restricted use of it for the determination of atomic 
 weights. Of Berzelius' work in this direction we shall 
 have occasion to speak later. 
 
 Just about this period in the history of the atomic 
 theory (1813), an Italian chemist, Avogadro, pointed out 
 that the whole of the differences between Dalton's and 
 Berzelius' attitudes towards the law of volumes would 
 disappear if chemistry would but admit into its philo- 
 sophy a new order of particles of a higher grade of 
 organisation than the atoms. These particles he called 
 molecules, and postulated that all the molecules of the 
 same substance are identically similar, and in general 
 consist of an assemblage of atoms, even in the case of 
 simple gases. Heretofore a radical distinction had 
 existed between the constitution of a simple, and that 
 of a compound, gas. A mass of oxygen was pictured 
 as an assemblage of atoms, each with perfect freedom, 
 and completely independent of its neighbours. In a 
 compound gas, such as hydrochloric acid, on the other 
 hand, the particles enjoying this individuality and 
 freedom were not single atoms else the gas had been 
 a mere mixture of hydrogen and chlorine but atom 
 complexes. Each atom of hydrogen kept perpetual 
 company with an atom of chlorine, the combination 
 forming a discrete and independent particle of the 
 gas. 
 
 Avogadro asserted that this distinction between the 
 simple and compound gases was unwarranted. He 
 regarded both hydrogen and chlorine as, in a sense, 
 
90 LAW AND THEOEY IN CHEMISTRY 
 
 compounds ; the one was to him hydride of hydrogen, 
 the other chloride of chlorine. The freely moving and 
 independent particle of hydrogen gas was not a single 
 atom, it was an atom complex. The only difference 
 between a simple and a compound gas is that the atom 
 complex of the former is made up of similar, that of 
 the latter of dissimilar, atoms. Having premised so 
 far, Avogadro then concluded that equal volumes of all 
 gases, simple and compound, contain under similar 
 conditions of temperature and pressure the same num- 
 ber of molecules. This conclusion is generally known 
 as Avogadro 's " hypothesis." ] As will be readily 
 noticed, it is nothing else than Gay-Lussac's conjecture 
 with one word changed molecules replaces atom. 
 
 In 1814, Ampere independently came to precisely 
 the same general conclusion with regard to the struc- 
 ture of gases, simple and compound. 
 
 It is easy to see that those instances of gaseous 
 combination wherewith Dalton combated Berzelius' 
 interpretation of the law of volumes involve no in- 
 consistencies when interpreted in the light of Avogadro 
 
 1 Avogadro's " hypothesis " is sometimes erroneously stated in the 
 form : All gaseous molecules under like conditions have the same size. 
 What is really meant is that all molecules have the same sized spheres 
 of action, which they occupy and dominate to the exclusion of other 
 molecules. A compact square of fifty soldiers armed with rifles would, 
 in a certain sense, dominate the same extent of country as twenty- 
 five soldiers armed with the same rifles, although the size of the 
 actual squares would be different in the two cases. It may here be 
 stated that Avogadro's generalisation is strictly true only for perfect 
 gases. The ratio of the number of molecules in equal volumes of 
 oxygen and hydrogen at ordinary temperatures and pressures is about 
 100,020: 100,000. 
 
THE ATOMIC THEORY 91 
 
 and Ampere's views. The volumetric relations of 
 hydrochloric acid gas merely prove that the molecules 
 of hydrogen and chlorine must contain at least two 
 atoms each. 1 
 
 Yet the times were not ripe for this great generali- 
 sation which to-day, under the title Avogadro's law, 
 stands the very foundation and framework of theoretical 
 chemistry. It attracted but little attention at the time 
 of its birth, and soon fell into an oblivion from which 
 
 1 The general introduction of these conceptions of the structure of 
 elementary gases at a later period in the history of chemistry, threw a 
 new light on the phenomena of substitution which for a time enjoyed 
 a special prominence. In the formation of a halogen derivative of a 
 hydrocarbon, it had up to date been necessary to ascribe different 
 roles to the admittedly perfectly similar atomic units of the halogen. 
 For instance, in the formation of monochlor-methane, it was held that 
 an atom of chlorine first replaced an atom of hydrogen in the hydro- 
 carbon 
 
 CH 4 + Cl = CH 3 C1 + H, 
 
 and then a second atom of chlorine combined with the liberated hydro- 
 gen atom 
 
 H + Cl = HC1. 
 
 In other words, substitution was regarded as a complex chemical change 
 taking place in two stages, the chlorine atom in each stage playing a 
 different chemical role. But when it was admitted that the smallest 
 portion of chlorine entering into chemical reaction is a molecule con- 
 taining two atoms, then halogen substitution was seen to be a pure 
 case of double decomposition, impossible of resolution into two con- 
 secutive stages, and recalling in its main features exactly the substi- 
 tution brought about by the action of compound bodies such as nitric 
 acid. 
 
 The two reactions 
 
 CH 3 H + C1.C1 = CH 3 C1 + HC1. 
 C 6 H 5 H + OH.NOo = C 6 H 5 N0 2 + HOH. 
 
 were seen to be quite analogous. It was no longer necessary to ascribe 
 different roles to the smallest individual particles of the halogens. 
 
92 LAW AND THEORY IN CHEMISTRY 
 
 it was rescued some forty years afterwards by Gerhardt 
 and Laurent. 
 
 In the year 1818, Dulong and Petit, experimenting 
 on the specific heats of the elements in a solid state, 
 discovered a most striking numerical relation between 
 these values and the atomic weights of the respective 
 elements. The numerical value of the atomic weight 
 of an element multiplied by the value of its specific 
 heat gave, in general, a constant product equal to 
 about 6'25. The limitation "in general" is advisedly 
 introduced, because certain of the current atomic 
 weights did not satisfy this relation. On the strength 
 of the comparatively large number of atomic weights 
 which did satisfy their law, Dulong and Petit pro- 
 nounced inaccurate all those which did not, thus 
 virtually asserting the universality of their law, and 
 establishing it as a powerful instrument for atomic 
 weight determinations. They recognised the arbitrari- 
 ness of the methods of determining atomic weight 
 then in vogue ; for these made the choice of one par- 
 ticular value out of a series, bearing a simple multiple 
 relationship to each other, a matter of, we might 
 almost say, individual taste. 
 
 In a small annex to a certain kitchen in the city of 
 Stockholm, Dulong and Petit's law was warmly wel- 
 comed. In this kitchen (which has been referred to as 
 one of the magnetic poles of the chemical world), assisted 
 by his cook and equipped with culinary utensils rather 
 than with what we now understand by apparatus, 
 worked one whose name, Johann Jacob Berzelius 
 
THE ATOMIC THEOEY 93 
 
 (1779-1848), is one of .the greatest on the honour rolls 
 of science. From the very beginning of his scientific 
 career, Berzelius interested himself chiefly in atomic 
 weight determinations. To convert his analytic results 
 into values for the atomic weights, he at first made use 
 of (1) certain rules recalling in their arbitrariness those 
 of Dalton, (2) the law of volumes already mentioned, 
 and (3) his so-called oxygen law. 1 Afterwards he 
 employed in addition to these (4) the law of Dulong 
 and Petit, and (5) the generalisation of Mitscherlich, 
 which has of late years proved itself the very reverse 
 of general. 
 
 At the outset, Mitscherlich believed that the correla- 
 tion between crystalline form and composition was 
 such, that a mere equality in number of the simple 
 atoms in the compound atoms of two substances, 
 necessitated an identity in the crystal forms, or an 
 isomorphism of the two substances ; and vice versa, that 
 isomorphism necessarily existed between substances 
 whose compound atoms were built up by the same 
 number of simple atoms. The discovery of poly- 
 morphism (the crystallisation of one and the same 
 substance in different forms), however, compelled the 
 
 1 Berzelius regarded salts as dual compounds of acid oxide or 
 negative constituent, and basic oxide or positive constituent. Sulphate 
 
 of soda was written * The oxygen law stated that in all the 
 
 salts of a given acid the amount of oxygen in the negative constituent 
 bore a constant ratio to the amount present in the positive con- 
 stituent. This law had been previously recognised by the German 
 chemist Richter, who, however, had expressed himself rather obscurely 
 on the point. 
 
94 LAW AND THEORY IN CHEMISTRY 
 
 admission that atomic complexity alone was not the 
 whole explanation of isomorphism, but that the arrange- 
 ment of atoms must be taken into account. Further, 
 the fact had to be admitted that in the majority of 
 cases the forms of crystals of similarly constituted 
 bodies are only approximately the same, not absolutely 
 identical. Indeed, absolutely perfect geometric iso- 
 morphism is only found in the cases of bodies crys- 
 tallising in the cubic system. Hence Mitscherlich's 
 generalisation in its final form was not the clear-cut 
 and unequivocal statement that it was in its original 
 enunciation. The final statement ran as follows : If 
 the compound atoms of two or more chemically analogous 
 bodies be composed of the same number of simple atoms 
 (no matter the nature of the latter), then the crystals 
 of these bodies will have identical or nearly identical 
 forms. 1 
 
 1 Recent research has adduced many exceptions to this elastic gene- 
 ralisation, both in its direct and converse forms. The three dinitro- 
 benzenes belong to the same chemical type, and have all the same 
 atomic complexity ; yet they have so little analogy of form that it 
 would be an obvious overstretching of the term to call them iso- 
 morphous. The exceptions to the generalisation in its converse form 
 are grouped together under the titles isogonous or homeomorphous 
 bodies. Such bodies, while differing greatly in chemical behaviour 
 and even in molecular complexity, are nevertheless isomorphous, e.g., 
 PbCl-2 and Sn(CH 3 ) 2 Cl 2 are isomorphous, as are also KHSO4 and 
 KAlSi 3 O 8 . It is clear throughout that the term isomorphous, which 
 was originally a definite and definable term, is now indefinite and 
 ^indefinable. What degree of similarity in the geometrical forms of 
 crystals is necessary in order that the crystals fall in the category, 
 isomorphous, is a question which chemists are undecided about. They 
 prefer to judge of isomorphism by a series of chemico-physical tests 
 rather than by purely crystallographic considerations. Two substances, 
 A and B, crystallise in the same system with nearly the same forms 
 
THE ATOMIC THEORY 95 
 
 So marvellously did Berzelius balance probabilities, 
 so carefully did he make use of analogy, so skilfully did 
 he manipulate his roughly improvised apparatus, that 
 his final table of atomic weights (1826), when reduced 
 to the hydrogen standard, 1 shows remarkable agree- 
 
 and angles. Are they isomorphous ? If they show the same cleavages, 
 similar thermal conductivities, similar etched figures, approximately 
 equal specific volumes ; if a crystal of A grows regularly in a solution 
 of B, or vice versa ; if a crystal of A causes crystallisation in a super- 
 saturated solution of B, or vice versd ; and if mixed solutions of A and 
 B give homogeneous mixed crystals, then the answer is most decidedly 
 yes. It is not a sufficient criterion of isomorphism that one only of 
 these conditions is satisfied, just as an element cannot in general be 
 classified as a metal or a non-metal from the investigation of a single 
 property. For a good account of isomorphism, regarded as a branch 
 of morphotropy (the general study of the inter-relation of chemical 
 composition and crystalline form apart from considerations of similarity 
 in the latter), see Hutchinson's article, "Isomorphism," in Watts 
 Dictionary of Chemistry, vol. iii. ; also Mendeleeff, Principles of 
 Chemistry, vol. ii. p. 7. 
 
 1 Berzelius regarded oxygen as the most important chemical element 
 the pole of chemistry. He therefore adopted the oxygen atom as 
 the basis of his atomic weight system, giving it the arbitrary value 100. 
 Modern chemistry has rather favoured Dalton's choice of the hydrogen 
 atom as standard atom with the arbitrary value 1. Quite recently, 
 however, the question of making oxygen the standard element again, 
 assigning it the arbitrary atomic weight 16, has been much discussed. 
 There is much to be said in favour of the return. The values of a 
 large majority of the atomic weights involve an accurate knowledge of 
 the atomic weight of oxygen. Unfortunately the determination of the 
 atomic weight of oxygen referred to H = 1 is an extremely difficult 
 chemical task, and new results differing often in the first decimal place 
 are continually demanding recognition. Every adoption of a new value 
 necessitates the alteration of the atomic weights of all those elements 
 whose oxy-compounds have furnished the necessary analytical data. 
 If O = 16 were universally adopted as the standard, the only atomic 
 weight that need be affected by new data for the composition of water 
 would be that of hydrogen ; and for all practical purposes the changes 
 it would undergo might with safety be overlooked and the value 1 
 steadily adhered to. A small change in the accepted value of the 
 
96 LAW AND THEOKY IN CHEMISTRY 
 
 ments with the values for the atomic weights current 
 to-day. To this statement we must, however, make 
 three well-defined exceptions, viz., potassium, silver, 
 and sodium. Lacking the data for the application of 
 the law of specific heat in these three cases, values 
 almost exactly twice too great were assigned to the 
 atomic weights of these elements. 
 
 I would emphasise the fact that Berzelius made use 
 of no one universally applicable guide in constructing 
 his system of atomic weights. He applied to the results 
 of analysis sometimes one, sometimes another, of the 
 five methods above given, and in cases of doubt he 
 selected with something akin to inspiration. One is 
 almost tempted to say that Berzelius was lucky. 
 
 Despite the great renown of Berzelius as an analytical 
 
 atomic weight of oxygen may involve quite a large change in the values 
 of other atomic weights, or it may involve none at all. The following 
 approximate values necessary for the determination of the atomic 
 weight of barium by Struve's method are instructive. Struve deduced 
 the atomic weight of barium from the ratio BaClo : BaSO 4 :: 100 : ] 12'1, 
 assuming values for the atomic weights of Cl, O, and S. 
 
 = 16 = 15-96 
 
 (1) Molecular weight of KC1 deduced from } ~,.~ ^.oo 
 
 reduction of KG 10 3 . . . . ( ' 4 
 
 (2) Atomic weight of Ag deduced from ratio ) in^.g 107 '4 
 
 Ag : KC1 \ 
 
 (3) Molecular weight of AgCl from the reduc- ) 149.9 142-5 
 
 tion of AgC10 3 . . . 1 
 
 (4) Atomic weight of Cl (= mol. wt. AgCl- | o^.^ 35.1 
 
 at. wt. Ag) i 
 
 (5) Atomic weight of S from ratio Ag : Ag 2 S0 4 31 '1 30 '9 
 
 (6) Atomic weight of Ba .... 135'6 132-66 
 
 An initial error in the atomic weight of oxygen of | per cent, is 
 multiplied up into a percentage error of 2*2 in the atomic weight of 
 barium as determined by Struve. 
 
THE ATOMIC THEORY 97 
 
 chemist of the highest rank, many refused to accept his 
 estimates of the atomic weights, not only because they 
 differed in many instances from those upheld by the 
 school of Dalton, but also on account of the discovery of 
 inaccuracies and inconsistencies inherent in the system 
 itself. 
 
 Using Berzelius' value for the atomic weight of 
 carbon, 12 '2, many anomalies had been noticed in the 
 results of organic analysis. The atomic weight of this 
 element was therefore redetermined by Dumas and 
 others, with the result that the Berzelian value was 
 found -2 too high. This discovery not only shook the 
 confidence of some of Berzelius' disciples in their master, 
 but was made by his opponents the occasion of heaping 
 all manner of sarcasm and unjust criticism on the great 
 Swedish chemist. 
 
 But perhaps Dumas' work on the vapour densities 
 of elements which are solid or liquid at ordinary tem- 
 peratures (1827) did more to bring the Berzelian system 
 of atomic weights into discredit than did the detection 
 of a If per cent, error in the atomic weight of carbon. 
 Berzelius had arrived at the values Hg = 200 and 
 P = 31 for the atomic weights of these elements 
 by a combination of some of the methods above 
 indicated. But Dumas found that the vapour of 
 mercury is only 100 times heavier than that of hydrogen 
 under the same temperature and pressure conditions ; 
 while the vapour of phosphorus is 62 times heavier 
 than that of hydrogen. Therefore if, as Berzelius 
 
 had maintained, equal volumes of gases contain equal 
 
 G 
 
98 LAW AND THEORY IN CHEMISTRY 
 
 numbers of atoms, then the atomic weights of mercury 
 and phosphorus must be respectively 100 and 62. If, 
 on the contrary, Berzelius' values were the true ones, 
 then his interpretation of the law of volumes an inter- 
 pretation lying at the very foundations of his atomic 
 weight system must necessarily lack generality. The 
 atomic weights of mercury and phosphorus being re- 
 spectively 200 and 31, a given volume of mercury 
 vapour can only contain half as many atoms as the 
 same volume of hydrogen ; while a given volume of 
 phosphorus must contain twice as many atoms as the 
 same volume of hydrogen. 
 
 The atomic theory now enters upon the most troub- 
 lous period of its career. So many different values 
 for the atomic weights of the elements were compet- 
 ing with each other for general acceptance, and so 
 vanishingly small did the chances of any universal 
 agreement on the subject appear, that it was proposed 
 to do away with the atomic theory and its attendant 
 uncertainties altogether, and to return to a system of 
 constants (with suitable notation) for the elements, 
 which constants, being simply the numerical expressions 
 of ascertained facts, could involve no doubt and admit 
 of no uncertainty. 
 
 This new system, which was to bring peace and 
 prosperity in its wake, was called the equivalent system, 
 and was pioneered by Wollaston. According to Wollas- 
 ton the symbol of an element was to represent that 
 mass thereof which combines with unit mass of hydrogen 
 a pure number expressing the result of an experi- 
 
THE ATOMIC THEORY 99 
 
 ment, nothing more. 1 The numerical value of this 
 mass he called the equivalent number, or simply, the 
 equivalent of the element. Just as those quantities 
 of acids which neutralise, i.e., combine with, a fixed 
 quantity of base are equivalent, so those quantities 
 of the elements which combine with a fixed mass of 
 hydrogen were also regarded as equivalent. 
 
 Yet this new system, seemingly so simple and un- 
 equivocal in its inception, soon had to encounter 
 difficulties as great as any that had ever beset the 
 atomic theory. Only a few of the elements could be 
 made to combine with hydrogen directly to form 
 hydrides ; how then were the equivalents of the 
 remaining elements to be determined ? One unit 
 mass of hydrogen combines with 35 J unit masses of 
 chlorine, and chlorine can be made to combine with 
 nearly all the elements that do not form hydrides. 
 Here then seemed a way out of the difficulty; the 
 equivalent of an element was taken to be that 
 mass thereof which combines with unit mass of 
 hydrogen, or 35J unit masses of chlorine. 2 But 
 unfortunately in cases where an element combines 
 
 1 As a matter of fact, Wollaston's standard of equivalency was not 
 unit mass of hydrogen, but ten unit masses of oxygen. This, however, 
 does not in any way affect the line of argument in the text (see note 
 1, p. 95). 
 
 2 It is to be observed that the equivalent of any element may be 
 determined as well by investigating the mass of hydrogen which a 
 given mass of the element replaces, as by finding the mass of hydrogen 
 with which a given mass of the element combines. For every combi- 
 nation of an element with chlorine, bromine, &c., may be regarded as 
 a substitution product of hydrochloric acid, hydrobromic acid, &c. 
 
ioo LAW AND THEORY IN CHEMISTRY 
 
 with both chlorine and hydrogen, the equivalent de- 
 duced from the chloride does not always coincide with 
 that deduced from the hydride. Again some elements, 
 notably carbon, form numerous compounds with hydro- 
 gen (methane, ethane, ethylene, &c., in the case of 
 carbon), and each of these compounds would give a 
 different value to the equivalent for carbon. Who was 
 to decide which particular compound was to be selected 
 for fixing the value in question? That an element 
 should be possessed of several equivalents seemed a 
 pure contradiction in terms. 
 
 Further complications arose when Gmelin and Gay- 
 Lussac attempted to make equivalent the formulas of 
 all compound bodies, including the most important 
 class of salts. 1 As long as the terms equivalent, or 
 equivalent weight, are restricted to acids and bases 
 among compound bodies, they have perfectly clear and 
 definite meanings, and are still employed in this con- 
 nection in modern volumetric analysis. But it is hard 
 to appreciate the application of the terms to salts. 
 However, this application was made in different ways 
 by different chemists, with the result that instead of 
 applying the equivalents of the elements deduced from 
 hydrides, chlorides, &c., to transform the results of 
 analysis of compounds into formulas for these com- 
 
 1 Here the idea of equivalency is somewhat changed. " AgO.S0 3 , 
 and NaO, 80s, are equivalent, not because they have equal powers of 
 displacement or combination with regard to any criterion, but because 
 they are the results of the union of bodies in such proportions that the 
 equal powers of combination of the constituent parts were satisfied in 
 the act of combining." 
 
THE ATOMIC THEORY 101 
 
 pounds, the problem was often reversed, and the 
 equivalents of the elements were deduced from pre- 
 sumed equivalences between salts. 1 The values obtained 
 in this way did not in general agree with those directly 
 obtained from analysis of hydrides, chlorides, &c. Thus 
 both the values 4f and 14 for the equivalent of nitrogen 
 found supporters ; while the equivalent of phosphorus 
 was either 10J, 15 4 5, or 31. 
 
 It is almost impossible for us now to appreciate 
 fully the reasoning of the chemists of this period. The 
 confusion of ideas which prevailed, 2 the gratuitous 
 
 1 See Wurtz, The A tomic Theory, p. 71, et scq. 
 
 2 Here it may be well to distinguish between the three terms com- 
 bining weight, equivalent, and atomic weight. The idea of combining 
 weights was associated with, and early recognised as a necessary 
 consequence of, the atomic theory (see note 1, p. 77). The atomic 
 weights were either numerically equal to, or whole multiples of, certain 
 characteristic numbers called combining weights that could be assigned 
 to the elements from analyses of their compounds, quite independently 
 of any assumptions or suppositions. Hence it would seem that the terms 
 combining weight and equivalent mean essentially the same thing ; 
 they do. The only difference between them is a chronological one. 
 The equivalent notation came after the atomic theory, and was in- 
 tended to be independent of it. It was an attempt to replace what up 
 to date had proved itself from a chemical standpoint an unsatisfactory 
 theory. Combining weights had their names changed to equivalents 
 when the atomic theory had been weighed and found wanting. The 
 reason for this change in nomenclature is not far to seek. It is an 
 epitomised history. The atomic weights of the Daltonian school were 
 in many instances numerically equal to the combining weights assigned 
 to the elements, and the two terms, though radically distinct, thus came 
 to be used more or less synonymously. Wollaston, in proposing the 
 name equivalent weights for the constants of his new system, was 
 simply desirous of avoiding a term which had become, quite wrongly, 
 more or less identified with a theory. The term equivalent was to 
 connote fact, and fact alone. 
 
 The difference between the terms atomic weight on the one hand, 
 
102 LAW AND THEORY IN CHEMISTRY 
 
 assumptions which flourished, and the reckless use of 
 the analogic method, combine to render its history one 
 of the most unsatisfactory and bewildering pieces of 
 modern scientific literature. Owing to the numerical 
 coincidences in many cases of the atomic weights and 
 the equivalents of the elements, the two terms, though 
 fundamentally so distinct in their connotations, came 
 to be used indifferently and synonymously. 1 
 
 It was Gerhardt and Laurent who, resuscitating the 
 long-eclipsed hypothesis of Avogadro, led the way out 
 of this confusion worse confounded. The equivalent 
 system presented to the mind of Gerhardt great in- 
 consistencies, which, in his opinion, could only be 
 
 and equivalent or combining weight on the other, is more fundamental 
 than the purely chronological one distinguishing equivalent from com- 
 bining weight. The term atomic weight implies a theory of the 
 structure of matter, the other two terms do not. The atomic weight 
 of chlorine is 35'5 ; its equivalent or combining weight is also 35*5. 
 The former statement calls up the following mental picture. Chlorine 
 gas is made up of a number of indivisible ultimate particles called 
 atoms equal among themselves, and each 35^ times as heavy as the 
 similar ultimate particles of which a mass of hydrogen consists. 
 Whereas the latter statement simply implies that a mass of chlorine 
 weighs 35^ times as much as the mass of hydrogen with which it 
 chemically combines. Though all idea of an equivalent notation is 
 now abandoned, the terms equivalent and combining weight are still 
 used, but now synonymously to denote the smallest mass of an element 
 that combines with unit mass of hydrogen, or 35^ unit masses of 
 chlorine, or with (approximately) 8 unit masses of oxygen. Indeed 
 it will be shown further on that the equivalent, as thus defined, really 
 determines the final value adopted for the atomic weight of an element. 
 1 Even as late as the year 1834 we find a committee of the British 
 Association for the Advancement of Science passing the following reso- 
 lution : "We are of opinion that it would save much confusion if 
 every chemist would always state explicitly the exact quantities which 
 he intends to represent by his symbols. " 
 
THE ATOMIC THEORY 103 
 
 removed by a partial return to the Berzelian system 
 of atomic weights. The inconsistencies that especially 
 appealed to Gerhardt were something of this nature. 
 The equivalents of carbonic anhydride and water being 
 represented as C0 2 and HO respectively (C = 6, = 8), 
 Gerhardt noticed that when these substances were pro- 
 duced in reactions involving organic bodies, they always 
 appeared in such relative quantities that the accepted 
 equivalents of the organic bodies could never be repre- 
 sented as giving rise to a single equivalent, but always 
 to 2, 4, or more equivalents of the oxides of carbon 
 and hydrogen. 
 
 Thus the interaction of acetic acid and sodium car- 
 bonate was represented as follows 
 
 C 8 H 8 8 + 2 NaC0 3 - C 8 H 6 8 Na 2 + 2 C0 2 + 2 HO. 
 
 It seemed incongruous and unnatural that organic 
 bodies should be so differently constituted from in- 
 organic ones as to be incapable of furnishing single 
 equivalents of water and carbonic anhydride when 
 undergoing chemical transformations. Gerhardt there- 
 fore proposed a return to the Berzelian values C = 12, 
 = 16, S = 32. Adopting these values, the difference 
 noted between organic and inorganic substances dis- 
 appears. The formulae for the equivalents of water, 
 sodium carbonate, and acetic acid become now H 2 0, 
 Na 2 C0 3 , and C 4 H 8 4 respectively. The formula for 
 the equivalent of carbonic anhydride still remains C0 2 , 
 the equivalent weight being, however, doubled. The 
 
104 LAW AND THEORY IN CHEMISTRY 
 
 equation representing the interaction between acetic 
 acid and sodium carbonate now takes the form 
 
 C 4 H 8 4 + Na 2 C0 3 = C 4 H 6 Na 2 4 + C0 2 + H 2 
 
 involving only single equivalents of carbonic anhydride 
 and water. 
 
 But there still remained in Gerhardt's opinion a 
 want of uniformity in the equivalent notation. He 
 noticed that the current formulae representing the 
 equivalent weights of compounds were not strictly 
 comparable, in that quantities of bodies correspond- 
 ing to their formulas expressed in grams, 1 say, did not 
 occupy the same volume when converted into the 
 gaseous state under similar conditions. Thus, C0 2 , 
 C 4 H 12 2 , and C 4 H 8 4 were the accepted formulas for 
 the equivalents of carbonic anhydride, alcohol, and 
 acetic acid respectively. 
 
 Volume in gaseous state (re- 
 
 TTI i 4. -m \^4- duced to and 760 mm.) of 
 Equivalent Weight. equivalent weight expres ' sed 
 
 in grams. 
 
 C0 2 .... 44 22-4 litres 
 
 C 4 H 12 2 ... 92 44-8 
 
 C 4 H 8 4 ... 120 44-8 
 
 Now Gerhardt proposed to make such alterations as 
 were necessary to reduce all formulae to a common gaso- 
 
 1 Of course the interpretation of formula) in grams is quite arbitrary, 
 but so long as we are ignorant of the absolute weight of the hydrogen 
 atom we cannot fix the absolute weights of compound atoms. What 
 we practically do is to make the conveniently simple and concrete 
 supposition (which we know to be far from true) that the hydrogen 
 atom weighs 1 gram, and the compound atoms of carbonic anhydride, 
 alcohol, and ether, 44 grams, 92 grams, and 120 grams respectively, 
 
THE ATOMIC THEORY 105 
 
 metric standard. Those masses of substances were to 
 be equivalent which occupy the same volume in the 
 gaseous state, and the formulae representing the com- 
 positions of these masses were to be the equivalent 
 formulas. It is evident that this reform of Gerhardt's 
 is nothing else than a return to Avogadro's " hypo- 
 thesis." Laurent clearly recognised this, and impressed 
 on Gerhardt the advisability of replacing the term 
 equivalent by molecule when it has reference to com- 
 pounds ; and by molecule or atom, as the case may be 
 when it is used in connection with elements. Hence- 
 forth Laurent's nomenclature will be observed. 
 
 Although Gerhardt returned to the Berzelian atomic 
 weights, yet it is interesting to remark that considera- 
 tions, not of an atomic structure of matter, but of an 
 equivalence, physical rather than chemical in its nature, 
 lay at the foundation of his reform, which ultimately 
 resulted in the complete abandonment of the equivalent 
 system so called. 
 
 The question as to which formulae he was to alter 
 next presented itself to Gerhardt. Was he to double 
 the formula of carbonic anhydride so as to make the 
 smallest particle, conserving all the properties of the 
 substance in mass, occupy the same volume as did 
 the accepted molecule of alcohol. Or was he to halve 
 the formula of alcohol so as to make its molecule 
 occupy the same volume as the accepted molecule of 
 carbonic anhydride ? To settle this question he had to 
 determine what volume a standard molecule, such as 
 that of hydrogen, occupies. From the volumetric com- 
 
io6 LAW AND THEOKY IN CHEMISTRY 
 
 position of hydrochloric acid gas, it follows that the 
 molecules of hydrogen and chlorine must each contain 
 an even number of atoms ; two at the least. That the 
 hydrogen molecule does not contain more than two 
 atoms is proved by the fact that hydrochloric acid is a 
 monobasic acid, and only under most exceptional con- 
 ditions forms sodium and potassium salts containing 
 hydrogen. 1 
 
 Hence the molecular weight of hydrogen is 2, and, 
 availing ourselves of the convenient supposition for 
 theoretical purposes of note 1, p. 104, we may provision- 
 ally say that it weighs 2 grams. But it is a well- 
 established constant that one litre of hydrogen under 
 standard temperature and pressure weighs '0896 
 grams. Therefore 2 grams of hydrogen will under 
 these conditions occupy 224 litres. 
 
 Hence it follows that carbonic acid has the right 
 formula, and that the formulae of alcohol, acetic acid, 
 and organic bodies generally must be halved. 2 
 
 1 Suppose for an instant that the molecule of hydrogen is tetratomic. 
 Then the formula of hydrochloric acid would be fL 2 C\x, and it would 
 necessarily be either a dibasic acid forming two sodium salts or a mono- 
 basic acid forming under all conditions salts of the type M'HCl^, 
 where M is a monovalent atom or radicle. 
 
 We have no good proof that the molecule of chlorine is not of higher 
 atomicity than 2. However, chlorine is chemically very analogous to 
 iodine, and it seems almost certain, from specific gravity determinations 
 carried out on the vapour of this body through a wide temperature 
 range, that its molecule does not contain more than 2 atoms. 
 
 2 The doubled formulae of organic bodies generally was due (1) to 
 the fact that these formulse were in many cases derived from analyses 
 of silver derivatives, and Berzelius' value for the atomic weight of 
 silver was twice too great ; (2) to the fact that certain views (embraced 
 under the title dualism) held at this period regarding the constitution 
 
THE ATOMIC THEORY 107 
 
 Gerhardt's proposed reform did not, however, meet 
 with general acceptance till after the brilliant work of 
 Williamson on the ethers. This research incontestably 
 showed by purely chemical reasoning the necessity for 
 halving the formula of alcohol in accordance with 
 Gerhardt's views, i.e., in accordance with the hypo- 
 thesis of Avogadro. 
 
 Starting with ethyl alcohol, Williamson hoped to pre- 
 pare therefrom an alcohol of higher molecular weight. 
 To this end he first treated ethyl alcohol with potassium 
 and then with ethyl iodide. Contrary to expectation, 
 the product was not an alcohol at all, but ordinary 
 ether C 4 H 10 0. This is inconsistent with the formula 
 C 4 H 12 2 for ethyl alcohol ; for if the latter body contains 
 per molecule twice as much oxygen as ether, then the 
 product of the above reactions ought to have contained 
 twice as much oxygen as ether, because there has 
 simply been replacement of H by C 2 H 5 . The only way 
 out of this difficulty appeared to be the representation of 
 alcohol by the halved formula C 2 H 6 0. 
 
 Alcohol. Ether. 
 
 2 H 5 I o C 2 H 5 I 
 
 Hj C 2 H 5 j 
 
 of bodies and the mechanism of chemical changes demanded the double 
 formulse for their expression. For details, see Meyer, History of 
 
 It follows that Gerhardt's halving of the molecular weights of acetic 
 acid, alcohol, &c., necessitated the halving of the Berzelian values for 
 the atomic weights of silver and the alkali metals. Unfortunately 
 Gerhardt carried his views to an extreme, and, without good warrant, 
 halved the atomic weights of twenty-three other elements, an error 
 which was afterwards pointed out and rectified by Cannizzaro, who 
 applied to the cases in question Dulong and Petit's law of specific heat, 
 
io8 LAW AND THEORY IN CHEMISTRY 
 
 True, a representation of this special formation of 
 ether could be given in terms of the doubled formula, 
 thus 
 
 2KI 
 
 But according to this mode of representation, a 
 mixture of methyl and ethyl ethers ought to result 
 if methyl iodide were used instead of ethyl iodide. 
 
 C 2 H 6 I 2 = C 4 H 10 + C 2 H 6 4 2 KI, 
 
 (ethyl ether) (methyl ether) 
 
 As a matter of fact, however, a single ether a so- 
 called mixed ether and not a mixture of two ethers 
 is the result ; and this can only be adequately repre- 
 sented in terms of the formula C 2 H 6 for alcohol. 
 
 (ethyl methyl ether) 
 
 Accordingly the molecular weight of alcohol was 
 halved, and thereupon the Gerhardt-Avogadro hypo- 
 thesis began to grow in general favour, and to acquire 
 that recognition which to-day classes it as the most 
 important instrument for the determination of atomic 
 weights. 
 
 Before demonstrating the use of the Gerhardt- 
 Avogadro generalisation in the determination of atomic 
 weights, we must first throw it into a more convenient 
 equational form. 
 
THE ATOMIC THEORY 
 
 109 
 
 X 
 
 
 Y 
 
 w grams 
 
 
 w' grams 
 
 n molecules 
 
 
 n molecules 
 
 Let H represent a volume of hydrogen weighing 1 
 gram. Let X and Y represent equal volumes of other 
 gases x and y, weighing respectively w and w' grams. 
 Let there be n molecules of hydrogen present in the 
 volume H ; then, assuming the truth of Avogadro's 
 generalisation, there will be n molecules of x in the 
 equal volume X and n of y in the equal volume Y. 
 
 Now the specific gravity 1 of x (referred to hydrogen 
 as standard) is obviously w. 
 
 And the specific gravity of y is w'. 
 
 Further, the weight of one molecule of x must 
 
 be grams, and the weight of one molecule of y 
 
 must be 
 
 Hence 
 
 n 
 
 Molecular weight of x _ n _ w _ Specific gravity of x. 
 Molecular weight of y w>_ w' Specific gravity of y. 
 
 In other words, Avogadro's generalisation may be re- 
 
 1 In chemistry, the term " vapour-density " is generally used when 
 " specific gravity of vapour " is really meant. Density is simply mass 
 per unit volume ; it is of two dimensions. Specific gravity is the ratio 
 of the density of a body to that of some standard substance. The 
 number expressing its value for any given substance is a pure number 
 and has therefore no dimensions. 
 
i io LAW AND THEORY IN CHEMISTRY 
 
 stated in the form: the molecular weights of gases 
 vary directly as their specific gravities. 1 
 
 Since x and y are any two gases whatever, let y 
 represent hydrogen. 
 Then- 
 Molecular weight of x Specific gravity of x. 
 Molecular weight of hydrogen Specific gravity of hydrogen. 
 
 Molecular weight of x __ Specific gravity of x. 
 ~~2~ 1 
 
 or, molecular weight of x = 2 x specific gravity of x. 
 
 If, as is frequently the case, air be taken as the stan- 
 dard of specific gravity, the above equation becomes 
 
 Molecular weight of x = 2 x 14-44 x specific gravity of x, 
 = 28 '88 x specific gravity of x, 
 
 air being, bulk for bulk, approximately 1444 times 
 heavier than hydrogen. 
 
 Having now thrown Avogadro's generalisation into a 
 numerical and equational form, we proceed to apply it 
 
 1 Avogadro's law admits of still another mode of expression. The 
 gaseous laws of Boyle and Charles are succinctly summed up in the 
 equation 
 
 pv = RT. 
 
 Where p is the pressure, v the volume, and T the absolute temperature 
 of a given mass of any gas, R is a constant depending (1) on the mass 
 of the gas taken, and (2) on the nature of the gas. However, if we 
 agree to apply the equation in all cases to masses of gases equal to 
 their Avogadrean molecular weights interpreted in grams, then R no 
 longer varies from gas to gas, but has the constant approximate value, 
 84,700 (the pressure being expressed in gravitation units grams per 
 square centimetre). 
 
THE ATOMIC THEORY 
 
 in 
 
 to the determination of an atomic weight, say, to the 
 determination of the atomic weight of oxygen. 
 
 I. 
 
 II. 
 
 III. 
 
 IV. 
 
 V. 
 
 Carbon monoxide . 
 
 42-86C : 57 '140 
 
 14 
 
 28 
 
 12 C : 16 
 
 Carbon dioxide . ' 27 '27 C : 72730 
 
 22 
 
 44 
 
 12 C : 32 
 
 Osmic oxide . . 74 '93 Os : 25 '07 
 
 127-6 
 
 255-2 
 
 191-3 Os : 64 
 
 Water . . . 1111 H : 88'890 
 
 9 
 
 18 
 
 2 H : 16 ! 
 
 Arsenious anhydride 
 
 75. 78 As: 24 '220 
 
 198 
 
 396 
 
 300-4 As : 96 
 
 Nitric oxide 
 
 46-67N : 53 -330 
 
 15 
 
 30 
 
 14 N : 16 
 
 Sulphuric anhydride 
 
 40'OOZ : 60-000 
 
 40 
 
 80 
 
 32 S : 48 
 
 &c. 
 
 &c. 
 
 &c. 
 
 &c. 
 
 &c. 
 
 Having selected several the larger the number the 
 better gaseous or gasifiable compounds of oxygen 
 (column i.), whose percentage compositions (column ii.) 
 are known to a fair degree of accuracy, we find their 
 approximate specific gravities in the gaseous state 1 
 (column iii.). From these values we deduce the approxi- 
 mate molecular weights 2 (column iv.) by means of one 
 
 1 For practical details see Muir and Carnegie, Practical Chemistry, 
 p. 121. 
 
 2 The definitions of molecular weight given in many of the text- 
 books are exceedingly vague, e.g. : 
 
 "The molecularweights are the weights of two volumes, for mole- 
 cules occupy two volumes if an atom of hydrogen occupies one." 
 
 "The molecular weight of a gaseous element or compound is a 
 number which tells the weight of two volumes of the gas, that is, the 
 weight of that volume of the gas which is equal to the volume occupied 
 by two parts by weight of hydrogen." 
 
 In the following definition, the attempt is made to avoid the 
 indefinite terms " volumes " and " parts by weight." 
 
 The molecular weight of any substance is that weight thereof 
 (expressed in terms of any unit whatever) which in the gaseous state 
 occupies the same volume as do two unit weights of hydrogen, the 
 same conditions of temperature and pressure and the same weight unit 
 being observed in both cases. 
 
ii2 LAW AND THEORY IN CHEMISTRY 
 
 or the other of the equations just developed. Then 
 we re-state the percentage compositions in such a way 
 as to give the molecular compositions (column v.). 
 The ratio 12 : 16 (column v.) is just the same thing 
 as the ratio 42'86 : 57-14 (column ii.), but 12 + 16 = 
 28 the molecular weight of carbon monoxide. And so 
 on for all the other bodies mentioned. 
 
 Now these molecular magnitudes have been deduced 
 from equations which are based on the supposition that 
 the smallest mass of hydrogen ever found in any 
 molecule (i.e., the atom of hydrogen), is numerically 
 represented by unity ; and a glance at column v. con- 
 vinces us that the smallest mass of oxygen found in 
 the molecule of any compound considered is, in terms 
 of the same mass unit, represented numerically and 
 approximately by 16. But the least amount of oxygen 
 that can by theory exist in a molecule is an atom. 
 Therefore the approximate atomic weight of oxygen is 
 16. 1 The method of procedure in the case of any other 
 element is exactly similar. 
 
 It still remains to show how the approximate atomic 
 weight assists us to a knowledge of the true atomic 
 weight, which now follows upon a very accurate deter- 
 mination of the equivalent of the element, i.e., the 
 
 1 We cannot positively and finally assert that the atomic weight of 
 oxygen is 16. We can only say that the probability of the approximate 
 value O 16 is almost infinitely great. A new substance might be 
 discovered whose molecule contained only 8i unit masses of oxygen 
 (H atom = 1 unit mass) ; and there are chemists who, desirous of 
 emphasising this vanishing possibility, speak of the maximum atomic 
 weight of oxygen being approximately equal to 16. 
 
THE ATOMIC THEORY 113 
 
 smallest mass of the element which combines with unit 
 mass of hydrogen, or with that mass of oxygen which 
 itself combines with unit mass of hydrogen. 
 
 In the case of oxygen the equivalent is not yet 
 definitely agreed upon (see note 1, p. 95). As a 
 probable result from all the recent elaborate deter- 
 minations of this most important constant, Ostwald 
 (LehrbucJi der Allgemeinen Chemie, p. 48) gives the 
 value 7'974 [H = 1]. Suppose W 7t grams of hydrogen 
 combine with W;,. grams of an element X, atomic 
 weight Z, to form a hydride whose molecular formula 
 
 W 
 
 is H m X n< Then is the equivalent of the element 
 
 X, and the ratio 
 
 Z x n W, 
 
 m W 
 
 or Z = equivalent number x 
 
 necessarily holds. In other words, the atomic weight 
 of any element is numerically equal to its equivalent 
 number multiplied by a fraction involving, as a rule, 
 only the lower digits. In general, m is a whole 
 multiple of n, so that the value of the atomic weight 
 becomes a whole multiple of the equivalent number, 
 or 
 
 Z = r x equivalent number, 
 where r = 1, 2, 3, &c. 
 
ii4 LAW AND THEOKY IN CHEMISTEY 
 
 In the case of oxygen, then, it follows that the 
 atomic weight may be either 
 
 7-974 
 
 or 15-948 = (7'974 x 2) 
 or 23-922 = (7'974 x 3) &c. 
 
 The approximate value 16, already arrived at by 
 the application of Avogadro's generalisation, enables 
 us to select without any hesitation from these possi- 
 bilities the value 15*948 as the interim atomic weight 
 of oxygen. 
 
 Until Kaoult's recent work, the constancy in the 
 volumes occupied by gaseous molecules (Avogadro's 
 generalisation) was the only known example of a 
 colligative property. 1 Kaoult established the fact that 
 
 1 Following Ostwald, we may divide all chemico-physical properties 
 into three classes ; additive, constitutive, and colligative. Weight is 
 the one and only true example of an additive property, for the weight 
 of a compound is exactly equal to the weights of the elements which 
 have reacted to form the compound. The majority of chemico-physical 
 properties which have been quantitatively investigated, belong to the 
 class of constitutive properties. Let A B D and ABC represent two 
 compounds of distinct types, and X', X" the respective values for 
 these bodies of some constitutive property, such as molecular volume 
 / _ molecular weighty Let ^ e and d repregent the atomic vo l u mes 
 \ specific gravity / 
 
 / atomic weight \ of ^ ^ Q and ^ respectively> in the free state> 
 \ specific gravity/ 
 
 Then X' and X" are not necessarily equal to a + b + d and a + b + c, 
 respectively ; for A B and D in compounds of the type A B D may 
 have volumes a' b' and d', while in compounds of the type ABC, 
 A and B may have still other values a" and b". Hence 
 
 X' = a' + b' + d' 
 
 X" = a" + b" + c". 
 
 It will be seen, however, that constitutive properties involve an additive 
 
THE ATOMIC THEORY 115 
 
 dilute solutions of indifferent bodies also exhibit well- 
 marked colligative properties. 
 
 A given quantity of solvent has its freezing point, or, 
 if volatile, its vapour pressure for a given temperature, 
 lowered a constant amount in each case, by the solution 
 therein of different substances, in quantities propor- 
 tional to their molecular weight. Just as the colligative 
 property expressed in Avogadro's generalisation gave 
 rise to a method for molecular weight determination 
 (and thence for atomic weight valuation), so Raoult's 
 results have in the hands of Beckman and others led to 
 new practical methods of great utility for the determina- 
 tion of molecular weights. These new methods are all 
 the more valuable in that they can be applied to bodies 
 which cannot be conveniently gasified, or cannot be 
 gasified without decomposition, i.e., to bodies which, 
 owing either to refractoriness or instability, fall outside 
 the pale of Avogadro's generalisation. 
 
 Since Van't Hoff showed that Raoult's empirical laws 
 can be derived by thermodynamical reasoning one from 
 the other, and both from a certain theory of solution 
 suggested by the phenomena of osmotic pressure, they 
 have been very widely and generally applied to the 
 determinations of molecular weights in cases unsuited 
 to the gaseous specific gravity method. 
 
 element ; and indeed, as soon as we know the values of any property 
 peculiar to the atoms in a given class of compounds, the evaluation of 
 constitutive properties for bodies of that class is a matter of pure addi- 
 tion. Avogadro's generalisation is the expression of a colligative pro- 
 perty. No matter what the composition or complexity of a gaseous 
 molecule may be, it will occupy under given conditions a given deter- 
 mined space. 
 
ii6 LAW AND THEOEY IN CHEMISTRY 
 
 Of the two methods due to Baoult, that founded on 
 the lowering of freezing point seems the most practical 
 and popular ; it has therefore received a special name, 
 being known as the cryoscopic method. For the details 
 of the practice and theory of these methods the reader 
 is referred to Ostwald's Solutions, translated from the 
 Lehrbuch der Allgemeinen Chemie. 
 
 It cannot be too strongly emphasised that chemists, 
 in applying to bodies in the solid and liquid states the 
 constants determined for the molecules of these bodies 
 in the gaseous state, do not thereby intentionally commit 
 themselves to any theory of the molecular structure of 
 solids and liquids. Chemical changes find adequate 
 interpretation in terms of the formulae derived from a 
 study of gases ; indeed, it seems as if the complex mole- 
 cules of solids and liquids, under the necessary condi- 
 tions for chemical change, and prior to such change, 
 break down into simpler molecules, comparable, if not 
 identical in magnitude, with the gaseous molecules. 1 
 In other words, it happens that the gaseous molecule 
 remains the chemically reacting unit in the solid or 
 liquid state, although the physical molecules (i.e., the 
 smallest masses of solids and liquids which exhibit all 
 the properties, physical and chemical, of the substances 
 
 1 Seeing that the unit of chemical activity of a solid or liquid is not 
 necessarily the physico-chemical molecule of that substance, but may be 
 only a dissociation product thereof, comparable in magnitude with the 
 true gaseous molecule, there are those who advocate the restriction 
 of the terms molecule and molecular weight to purely gaseous pheno- 
 mena, using the non-committal terms, reacting unit and reacting 
 weight, in describing the interactions of solids and liquids. 
 
THE ATOMIC THEORY 117 
 
 peculiar to their solid or liquid states), are almost 
 certainly polymers of the gaseous molecules. 
 
 If MN be the formula of the gaseous molecule, then 
 (M^Na;) and (M^N^) represent the physical molecules of 
 the same substance in the liquid and solid states re- 
 spectively ; x and y being whole numbers, and y in all 
 probability greater than x. 
 
 It may be that the time is coming when we shall ex- 
 press more in our equations than we do at present, this 
 increased information involving the true molecular 
 magnitudes of the solids and liquids (or dissolved 
 substances) interacting, i.e., the absolute values of x 
 and y. 
 
 That this polymerisation, or condensation of the 
 simpler into more complex molecules, does often take 
 place as a substance passes from the gaseous into the 
 liquid state is certain, if we are justified in the universal 
 application to vapours of Avogadro's generalisation. 
 With many substances it makes itself evident even 
 before the liquid state is reached. At high tempera- 
 tures and low pressures gaseous nitric peroxide has 
 the triatomic molecule N0 2 ; at lower temperatures 
 and higher pressures these simpler molecules combine 
 in pairs to form hexatomic molecules N 2 4 . Again, 
 acetic acid vapour at high temperatures is made up of 
 molecules C 2 H 4 2 , while at temperatures near its con- 
 densing point the molecules have the more complex 
 formula C 4 H 8 O 4 . Indeed in many cases the specific 
 gravities of the vapours of substances gradually increase 
 as the temperature of the boiling-point is approached, 
 
u8 LAW AND THEORY IN CHEMISTRY 
 
 indicating a gradual increase in complexity of the 
 molecules. 1 
 
 Now the gaseous and liquid states are really continu- 
 ous, despite the apparent discontinuity which appears 
 in the majority of cases under ordinary conditions of 
 temperature and pressure. Hence we are led to the 
 conclusion that the molecular condensation which ap- 
 pears in some cases as the liquid state is approached 
 (but in other cases is so delayed as not to manifest 
 itself until the upper limiting temperature of the liquid 
 state has been actually passed) will continue to take 
 place in increasing degree as the temperature is lowered 
 after the liquid state is reached. 
 
 That this conclusion is just has been verified in 
 
 1 The variability in the values of specific gravity determinations 
 of vapours at different temperatures is well manifested in the case of 
 the chlorides of aluminium and iron, and in this connection has caused 
 much discussion. Griinewald and Meyer hold that the specific gravity 
 of ferric chloride does not become approximately constant for any con- 
 siderable temperature interval until a temperature of over 700 has 
 been reached, when the vapour consists of molecules having the composi- 
 tion FeCla. On the other hand, Nilson and Pettersson maintain that the 
 specific gravity of the vapour is constant for quite a large temperature 
 interval, with 321 as its lower limit. Their experiments led to the 
 molecular formula Fe 2 Cl6. Again, while Friedel and Crafts upheld the 
 formula A1 2 C1 6 for aluminic chloride, Nilson and Pettersson are led by 
 their experiments to support the formula A1C1 3 . It seems necessary 
 to admit that in both cases molecules of the types M 2 C1 6 and MCl;j 
 exist ; an admission which surrounds the expression' "molecular weight 
 of a gas " with indefiniteness. For all purely equational purposes it 
 matters little which formula we use ; but the true molecular formula, 
 if indeed we can speak of a true molecular formula in these cases, is a 
 point of great moment regarded from the standpoint of the valency 
 theory (see Chapter VI.). Obviously the whole discussion turns on the 
 difficulty of proving the existence and nature of the decompositions 
 and dissociations taking place in a gas at high temperatures. 
 
THE ATOMIC THEORY 119 
 
 several cases by the application of Raoult's cryoscopic 
 method of molecular weight determination. 1 Yet it must 
 be remembered that the cryoscopic method can only be 
 applied to fairly dilute solutions, and it is almost cer- 
 tain that in such solutions the molecule of a dissolved 
 solid or liquid is smaller than it would be in the homo- 
 geneous solid or liquid state. The mere act of solution 
 often dissociates the molecules of the solid or liquid, 
 to such an extent that the dissociated molecules are of 
 the same magnitude as the gaseous molecules ; some- 
 times, however, the dissociation does not proceed so far, 
 the extent of the dissociation being dependent on the 
 nature of the indifferent solvent used. Acetic acid, 
 used as a solvent, seems to have a great dissociating 
 tendency ; hydrocarbon solvents, on the other hand, 
 manifest this property in a much less pronounced 
 degree. 2 In short the cryoscopic method can only 
 
 1 Measurements of other physical quantities also point to the exist- 
 ence of complex liquid molecules. Eb'tvos finds it impossible to explain 
 certain observations of his on surface tension, unless it be admitted 
 that molecules of liquids are from case to case more complex in vary- 
 ing degree than the molecules of the same substances in the gaseous 
 state. Recently Ramsay and Shields have come to similar general 
 conclusions. See Chemical Society's Journal, Ixiii. p. 1089. 
 
 2 Because certain liquids (e.g., acetic ether) possess in all solvents the 
 same molecular weight as they do in the gaseous state, it has been 
 suggested that the gaseous molecules of such bodies do not polymerise 
 when liquefaction takes place. This seems quite a misleading sug- 
 gestion. It is not probable that the dissociation tendency is a function 
 only of the solvent ; the greater or less stability of the molecular 
 complex of the dissolved substance itself must also be a factor. And 
 surely it is not improbable that in some bodies the molecular com- 
 plexes peculiar to the liquid state may be held together so loosely that 
 they break down even under the influence of comparatively inert 
 hydrocarbon solvents. 
 
120 LAW AND THEORY IN CHEMISTRY 
 
 furnish us with a minimum value for the molecular 
 weight of a solid or liquid body. 
 
 Of the true molecular weight of solids nothing is 
 known. It is supposed that the polymerisation of the 
 molecules in the solid state is greater than it is in the 
 liquid state that the molecule of ice is heavier than 
 the molecule of water. Perhaps one and the same solid 
 substance can have different weighted molecules accord- 
 ing to the conditions of its formation and existence, 
 and with this difference may, perhaps, be associated 
 the phenomena of allotropism (or physical isomerism) 
 and polymorphism (see p. 93, and Chap. VI.). 
 
CHAPTER V. 
 
 THE CLASSIFICATION OF COMPOUNDS. ACIDS, 
 BASES, SALTS. 
 
 BEFORE dealing with the more important classes of 
 compounds, it is advisable to preface a few remarks 
 on a classification of the elements which, in spite of 
 indefiniteness and artificiality, still enjoys a greater or 
 less degree of popularity. 
 
 At one time it was believed that chemical action 
 is essentially electrical in its nature, and that every 
 chemical change is the result of the play of the stronger 
 or weaker attractions of oppositely electrified atoms or 
 groups of atoms. Berzelius especially developed this 
 view of the electrical nature of chemical affinity. He 
 believed each atom to be electrically bipolar each 
 atom to have a definite charge of positive, and a 
 definite charge of negative, electricity. 1 Since these 
 charges were in general supposed to be unequal in 
 amount, the charge present in greater amount gave a 
 more or less positive or negative character to the atom 
 as a whole. In accordance with the results of experi- 
 ments on the behaviour of several compounds when 
 
 1 If Berzelius had expressed himself in terms of electric potential, 
 the necessity for assuming a bipolar electrical distribution in the atom 
 would have been obviated. 
 
122 LAW AND THEORY IN CHEMISTRY 
 
 electrolysed (i.e., when decomposed by the electric 
 current), Berzelius arranged the elements in an electro- 
 chemical series such that the element with the most 
 negative atoms stood at the head, and the element 
 with the most positive atoms at the bottom, of the 
 series, any intermediate element being more negative 
 than those below it in the series, and more positive 
 than those above it. 
 
 It was soon recognised that the elements towards 
 the electro-negative end of the series thus constructed 
 had properties in common which were different from 
 those common to the elements towards the electro- 
 positive end of the series. The electro-negative ele- 
 ments are, as a rule, of comparatively low specific 
 gravity, bad conductors of heat and electricity, and 
 more or less transparent. Their oxides are for the 
 most part readily soluble in water, and produce solu- 
 tions with a sour taste and great solvent and corrosive 
 powers. The elements themselves show an aptitude 
 for combining with hydrogen to produce hydrides and 
 decompose water, combining with its hydrogen and 
 setting free its oxygen. 
 
 The electro-positive elements, on the other hand, 
 have as a class relatively high specific gravities, and 
 possess a peculiar lustre. They are good conductors 
 of heat and electricity, and are translucent only when 
 reduced to very thin layers. Their oxides are for the 
 most part insoluble in water, but they have the power 
 of reacting with the solutions of the oxides of the 
 electro-negative elements so as to destroy all the char- 
 
THE CLASSIFICATION OF COMPOUNDS 123 
 
 acteristic properties of the latter. Further, the electro- 
 positive elements do not readily form compounds with 
 hydrogen, and decompose water, combining with its 
 oxygen and liberating the hydrogen. 
 
 It seemed good therefore to divide the elements 
 into two classes. 
 
 (1) The electro-negative elements or the non-metals. 1 
 
 (2) The electro-positive elements or the metals. 
 
 If this classification enabled us to make anything 
 approaching to a satisfactory definition of acids, bases, 
 and salts, it might despite its indefiniteness 2 be re- 
 tained. But we shall show further on that it does not 
 possess this much to be desired merit, and may also 
 remark here that the only satisfactory classification of 
 the elements that afforded by the periodic law does 
 not favour any such simple and fundamental division as 
 is embraced in the terms metal and non-metal. 
 
 Before passing on to acids, bases, and salts, it is also 
 
 1 By some the word " metalloid " is used in preference to " non- 
 metal." In view of the significance of the termination oid as 
 generally used, this preference is scarcely justified. 
 
 2 This indefiniteness is a necessary accompaniment of all classi- 
 fications which are not based on one definite criterion. A number 
 of properties are currently regarded as metallic, a number of other 
 properties as non-metallic. These properties in each case are not con- 
 sidered of equal value inter se in deciding as to the metallic or non- 
 metallic character of the element possessing them, but yet there exists 
 no authoritative conventional scale of values. An element as a rule 
 possesses properties falling under each category, and the non-metallic 
 properties have to be weighed against the metallic ones in the un- 
 adjusted balances of private opinion. In the absence of all convention 
 it is impossible to give a final and indisputable answer to the question, 
 " Is tellurium a metal or a non-metal ? " 
 
124 LAW AND THEORY IN CHEMISTRY 
 
 necessary to say a few words on a thoroughly arbitrary 
 classification of compounds which is in vogue. All 
 compounds, by convention, fall under one or other of 
 the two headings organic and inorganic; and hand 
 in hand with this classification goes a division of the 
 science at large into two branches organic and in- 
 organic chemistry. But "carbon compounds" and 
 " the chemistry of carbon compounds " are undoubtedly 
 much more appropriate titles than "organic compounds" 
 and "organic chemistry" respectively. For compara- 
 tively very few of the bodies treated in this main 
 subdivision of compounds are, as the latter titles seem 
 to imply, products of organisms, i.e., of life ; while, on 
 the other hand, some substances included among in- 
 organic compounds are products of vital metabolism. 
 
 Since Lemery's time (1645-1715) up till the be- 
 ginning of this century, chemists were possessed of the 
 idea that the products of organisms most of which 
 are compounds of carbon could not be made artifi- 
 cially, but that their production absolutely demanded 
 the intermediacy of life the operation of the so-called 
 vital force. "Operations of chymistry fall short of 
 vital force; no chymist can make milk or blood of 
 grass." Further, it was the custom to ascribe the 
 variable results obtained in analyses of animal and 
 vegetable products to the fact that such bodies did 
 not conform to the fundamental law of chemistry (law 
 of fixity of composition), and not to the difficulties 
 attendant on the analytical methods and the purifica- 
 tion of complex carbon compounds. Hence a very 
 
THE CLASSIFICATION OF COMPOUNDS 125 
 
 sharp distinction was made between organic and in- 
 organic or mineral compounds. 
 
 But the distinction is now recognised as existing 
 for convenience only. 1 Exactly the same laws hold 
 in the case of carbon compounds as in the realm of 
 mineral or inorganic chemistry, and many of those 
 carbon compounds which are products of the life of 
 organisms have been synthesised independently of the 
 living laboratory, i.e., of vital force so called. Thus 
 there exists no better reason than convenience to be 
 adduced for this primary division of compounds, and 
 no valid excuse exists for retaining the effete and mis- 
 leading term "organic compounds" in preference to 
 the more fitting title " carbon compounds." The com- 
 pounds of nitrogen are increasing so rapidly, that 
 before long it may be advisable to study them as a 
 class apart, just as has been the case with carbon 
 compounds. 
 
 In discussing acids, bases, and salts, it will be only 
 too apparent that we are unable to give satisfactory 
 definitions of these classes of compounds, and farther, 
 that any of the attempts at definition that may be 
 advanced are not independent one of another. In 
 other words, we cannot attempt to define one of these 
 classes of compounds without implying or assuming 
 the definitions of the other classes. In physics all 
 definitions can be expressed quite independently of one 
 
 1 A sign of the extremely arbitrary nature of this classification is 
 seen in the fact that the oxides of carbon, together with the bodies 
 they form by direct combination, are included in inorganic chemistry. 
 
126 LAW AND THEORY IN CHEMISTRY 
 
 another in terms of powers of three fundamental units, 
 but in chemistry we have no approach to such definition. 
 
 Acids. Probably the first acid known to the ancients, 
 certainly the one best known to them, was vinegar. 
 Regarding the solvent powers of this comparatively 
 feeble acid, most exaggerated notions were entertained ; 
 witness the belief that Hannibal therewith etched a 
 passage over the Alps for his army. The alchemists, 
 however, knew most of the mineral acids. Geber (eighth 
 century) describes the preparation of aquafortis (nitric 
 acid) and also of aqua regia, which he obtained by 
 dissolving sal ammoniac in aqua fortis. Geber also 
 knew of oil of vitriol. The iatro-chemist Valentine 
 (beginning of the sixteenth century) first made spiritus 
 salis (hydrochloric acid), and showed that aqua regia 
 can be made by mixing this new spiritus with aqua 
 fortis. Later iatro-chemists, Libavius and Glauber, 
 contributed much to an increased knowledge and more 
 extended use of the acids by introducing improved 
 methods of preparation. 
 
 But as yet no attempt at the classification of bodies 
 by their properties had been made, and the generic 
 term acid had not arisen. 
 
 Boyle was the first to group together into one class 
 all substances which have the following properties : 
 (1) a sour taste, (2) a great solvent action, (3) the 
 power of precipitating sulphur from alkaline solutions 
 of this substance, (4) a reddening action on many 
 vegetable blues such as litmus, (5) the power of acting 
 on wood ashes to produce substances without either 
 
THE CLASSIFICATION OF COMPOUNDS 127 
 
 the astringent solvent properties of acids or the soapy 
 cleansing properties of solutions of wood ashes. Bodies 
 possessing these properties were called collectively acids. 
 Even down to the beginning of the eighteenth century 
 most chemists were satisfied with the explanation that 
 all substances possessing these properties did so in 
 virtue of a greater or less amount of a common con- 
 stituent which was called the "primordial acid." Such 
 a vague explanation did not however satisfy Lavoisier, 
 who renounced all alchemistic fancies of "primordial 
 acids" and "principles of acidity." He thought, as 
 we have seen, that oxygen was the acid producer, and 
 that therefore every acid owes its acidity in some way 
 to the oxygen which it necessarily contains. It must 
 be here emphasised that Lavoisier did not regard the 
 elements of water, and therefore hydrogen, as necessary 
 constituents of acids. His acids were mere dissemi- 
 nations in water of our acid anhydrides (or acidic 
 oxides) binary compounds of some non-metal with 
 oxygen. Thus to Lavoisier sulphuric acid was S0 3 , 
 and not H 2 0, S0 3 , or H 2 S(V 
 
 Up to the year 1787 all the substances known to 
 
 1 It is now believed that after their solution in water, the acid 
 anhydrides as such lose their identity ; a new class of bodies called 
 hydroxides being formed. These bodies contain as proximate parts of 
 their molecules one or more hydroxyl (OH) radicles. Thus when S0 3 
 reacts with H 2 O, a rearrangement of atoms is believed to accom- 
 pany combination, so that the resulting compound is represented as 
 S0 2 (OH) 2 (sulphuryl di-hydroxide), and not as SO 3 , H 2 O (hydrated 
 sulphuric oxide). Similar remarks apply to the solutions of the oxides 
 of certain metals known as the alkaline oxides, and the alkaline -earth 
 oxides. Thus potassium oxide K 2 O reacts with water in the process 
 of dissolving in it to form potassium hydroxide K(OH). 
 
128 LAW AND THEORY IN CHEMISTRY 
 
 have the properties of acids as laid down by Boyle 
 were products of the interaction of water with the 
 oxides of non-metals ; hence the oxides of the 
 non-metals were called acid (acidic) oxides. In 
 the year just mentioned Berthollet took up the in- 
 vestigation of prussic acid (discovered some three or 
 four years previously by Scheele), and came to the 
 conclusion that it had true acid properties but yet was 
 entirely free from oxygen. In 1796 he came to the 
 same conclusion for sulphuretted hydrogen, or sul- 
 phydric acid, as it may be called. Lavoisier's reputa- 
 tion was, however, more mighty than Berthollet's facts, 
 and as regards acids, matters remained in statu quo 
 until Davy in 1810 clearly proved that hydrochloric 
 acid does not contain any oxygen, but is a compound 
 of hydrogen with an element chlorine. Further, in 
 1813, hydriodic acid was shown to be altogether free 
 from oxygen, and it was remarked that iodic anhy- 
 dride shows no acid properties until it is dissolved in 
 water, i.e., until the element hydrogen is introduced. 
 Davy then began to recognise that there is no one 
 particular acid-forming element. He noticed that 
 hydrogen is a universal constituent of acids, but he 
 did not rush over to the converse and say that all 
 hydrogenised bodies are necessarily acids. On the 
 contrary, he regarded the existence of acid properties 
 as depending chiefly on the other elements combined 
 with the hydrogen, and not on the hydrogen itself. 
 The result of Davy's work was that acids came to 
 be classed as hydracids (acids free from oxygen) and 
 
THE CLASSIFICATION OF COMPOUNDS 129 
 
 oxyacids (acids formed from acidic oxides). Shortly 
 afterwards Liebig came to the same conclusions as 
 Davy, and defined acids as particular compounds of 
 hydrogen in which the latter can be replaced by 
 metals. 
 
 The growth of our knowledge since Davy's time 
 has not brought us a satisfactory expression of any 
 generalisation as to the particular compounds of 
 hydrogen enjoying this property. We are in fact 
 bound to admit that an acid belongs to that category 
 of things which are perfectly conceivable but indefin- 
 able. With Boyle and Liebig we can enumerate the 
 properties which we agree shall be connoted by the 
 term acid we can agree as to what are to be regarded 
 as acidic functions, but that is all. The difficulty 
 remains that several bodies possessing some, or may 
 be all, of these properties are, for other reasons, not 
 regarded as acids. We have no one property which we 
 can use as an absolute criterion of acidity. Thus the 
 salt copper sulphate turns blue litmus red, and causes 
 an effervescence of carbon dioxide when mixed with 
 a soluble carbonate. Bisulphate of sodium turns blue 
 litmus red, causes effervescence with soluble carbonate, 
 and contains hydrogen which is replaceable by metals ; 
 yet on account of the method of its formation it is not 
 regarded as an acid. It seems hopeless then to attempt 
 any definition of the term acid in terms of chemical 
 properties. 
 
 If we attempt a definition from the point of view 
 
 of composition rather than properties, we meet with 
 
 I 
 
130 LAW AND THEORY IN CHEMISTRY 
 
 equally great difficulties. As Davy and Liebig saw, 
 all bodies regarded as acids contain hydrogen, but 
 all hydrogenised bodies are not acids. Evidently the 
 hydrogen must be combined with certain elements, 
 and in a definite manner, to produce an acid. Let 
 us try to generalise in this direction. The oxides 
 of the non-metals react with water to produce 
 hydroxides which are called acids (see note, p. 127). 
 Therefore we might attempt to define acids as the 
 hydroxides of the non-metals [Cl(OH)] or of non-met- 
 allic groups [S0 2 (OH)J. iBut the bodies HC1, HBr, HI, 
 &c., are called acids. Hence we must enlarge our 
 definition to include these hydracids. Acids are the 
 hydrides and hydroxides of non-metallic elements. 
 But this is not a successful definition ; for nitrogen 
 and phosphorus are generally regarded as non-metals 
 by those who employ the latter term, yet their hydrides 
 are not acidic. 1 Again, the bodies H 2 PtCl 6 and 
 H 4 FeC 6 N 6 are acids, but neither PtCl 6 nor FeC 6 N 6 
 can be called groups of non- metallic elements. It 
 seems hopeless then, even had the title non-metal 
 a strict connotation, to attempt to satisfactorily define 
 an acid in terms of its qualitative composition. We 
 cannot define an inorganic acid in terms of its pro- 
 perties alone, nor in terms of its composition alone, 
 nor in terms of both together; composition, properties, 
 and synthetical history must all be taken into account 
 in deciding a body's claim to rank as an acid. 
 
 1 Ammonia nevertheless contains hydrogen directly replaceable by 
 the metals K, Na. 
 
THE CLASSIFICATION OF COMPOUNDS 131 
 
 Acids are classified according to their strengths or 
 affinities, and also according to their basicities. The 
 former classification will be taken up in the chapter on 
 chemical equilibrium. The latter classification is due to 
 Liebig, who distinguished acids as mono-, di-, tri-, tetra- 
 basic, &c., according as they possessed one, two, three, 
 four, or more atoms of hydrogen per molecule replace- 
 able by metals. It is not always true that the replace- 
 able hydrogen atoms are coequal in number with the 
 total hydrogen atoms present in the molecule. In cases 
 where this equality exists the basicity of the acids can 
 at once be determined from their formulae ; but the 
 equality must have been proved by experiment to exist. 
 The experimental investigation of basicity takes the 
 form of a determination of the maximum number of 
 stable potassium or sodium salts the acids can form. 1 
 
 Bases. From early times it had been noticed that the 
 extracts of the ashes of burnt plants possessed certain 
 characteristic properties. For example, these extracts 
 have great cleansing power, apparently dissolving 
 grease and fat ; they are soapy to the touch, and have 
 the property of restoring to their normal colours vege- 
 table blues which have been reddened by acids. The 
 active principle of these extracts was called an alkali, 
 this title first appearing in Geber's writings, and liter- 
 ally meaning the ash. 2 Soon it was discovered that a 
 
 1 For practical details see Muir and Carnegie, Practical Chemistry, 
 p. 38 et seq. 
 
 2 The active principle of wood ashes is in reality potassium carbonate. 
 This is no longer classed as an alkali ; it is a salt with alkaline pro- 
 perties. 
 
132 LAW AND THEOKY IN CHEMISTRY 
 
 solution of the highly volatile ammonia or hartshorn 
 also possessed the above properties. Hence a distinc- 
 tion was made between fixed and volatile alkalis. It 
 was Stahl who first noticed that, in addition to the 
 above enumerated properties, alkalis have the power of 
 reacting with acids to produce indifferent substances. 
 It was as if the properties of acids and alkalis were 
 diametrically opposed, and in combination mutual 
 neutralisation took place, just as would be the case 
 with two equal and opposite magnetic poles a sort of 
 cancelling out of positive and negative. 
 
 So far the substances which had the power of thus 
 neutralising acids were all soluble in water. Sometime 
 later it was found that certain almost insoluble sub- 
 stances which are very fixed in the fire, i.e., do not 
 melt nor change in any way, have the same property. 
 These were called the earths, and Rouelle in 1744 
 combined alkalis and earths into the one class base, 
 defining a base as any substance which combines with 
 and neutralises an acid. 
 
 It is almost impossible to give a concise and 
 exact definition of base as that term is now applied. 
 Speaking loosely, the oxides of the metals, and the 
 hydroxides they form when they are made to combine 
 directly or indirectly with water, are regarded as bases ; 
 hence metallic oxides are often spoken of as basic 
 oxides. May we not then define a base as the oxide 
 or hydroxide of a metal ? Such a definition does not 
 include hydrides of certain non-metals (ammonia NH 3 , 
 phosphine PH 3 , &c.), which are established as bases ; 
 
THE CLASSIFICATION OF COMPOUNDS 133 
 
 and, further, it does not indicate that the higher oxides 
 of some metals have acidic functions. We are equally 
 unsuccessful when we try to frame a definition in 
 terms of properties. All bases do not react with acids 
 so as to neutralise the latter, i.e., completely destroy 
 their characteristic properties, as Rouelle, to whom 
 only the strongest bases were known, believed. The 
 solution obtained when sulphuric acid is treated with 
 an excess of the base copper oxide still possesses acid 
 characters. Again, while some bases act on vegetable 
 colouring matters, others do not. Nor can bases be 
 said to be bodies which in reacting with acids cause a 
 replacement of the hydrogen of the acids by metals, for 
 the bases N 2 H 4 , NH 3 , PH 3 , &c., do not contain metals, 
 and the products of their interaction with acids are 
 of the nature of additive, not substituted, compounds. 
 True, the definition, " A base is a substance which 
 reacts with acids to produce a salt and water at most," 
 holds good, but it is a definition in terms of acids which 
 cannot be defined, and involving a new class of bodies, 
 salts which themselves require definition. Hence we 
 must conclude that while it is possible to lay down 
 what we mean by basic characters or functions, it is 
 impossible to define bases ; for many bodies which are 
 not regarded as bases exhibit basic characters, and 
 many bases exhibit only a few of what are all admittedly 
 basic functions. 
 
 A rough classification of the more important inor- 
 ganic bases follows. The table clearly "shows the true 
 relation of the terms alkali and base. In some elemen- 
 
134 
 
 LAW AND THEORY IN CHEMISTRY 
 
 tary text-books the terms are used as if they were 
 synonymous. This is not the case. An alkali is a 
 particular kind of hydroxide which is a particular kind 
 of base. Every alkali is a base, but every base is not 
 an alkali. 
 
 Alkaline oxides (K,0, Na 2 0, 
 Eb 2 O, Cs 2 0, KbaO, Li 2 0) 
 form very" soluble hydroxi- 
 des with marked basic char- 
 acters called alkalis. 
 
 Alkaline earth oxides (CaO, 
 SrO, BaO) form slightly 
 soluble hydroxides with 
 less pronounced bisic char- 
 acters than the alkalis. 
 
 The lower oxides of the re- 
 maining metals fall under 
 this heading. The hy- 
 droxides of these oxides 
 are formed by adding an 
 alkali, or alkaline earth 
 hydroxide, to solutions of 
 
 . the oxides in acids. 
 
 Oxides and hy- 
 droxides of 
 the metals. 
 
 Oxides which directly 
 react with water 
 to produce hy- 
 droxides. 
 
 Oxides which do not 
 directly react with 
 water. 
 
 Hydrides of 
 certain non- 
 metals, and 
 their deriva- 
 tives. 
 
 > Ammonia NET 3 , amidogen NoH 4 , hydroxylamine 
 NH 2 OH, phosphine PH 3 , &c." 
 
 However convenient classifications of oxides into 
 acidic, basic, intermediate, and indifferent varieties 
 may be for an elementary presentment of chemistry, a 
 glance from the vantage-ground of. facts which are not 
 usually referred to in elementary courses, but which 
 are none the less facts on that account, shows such 
 classifications to be imperfect and arbitrary to a degree. 
 New reactions and new substances are continually 
 being discovered, which testify to the fact that oxides 
 generally are capable of exhibiting under one set of 
 conditions what are currently accepted as acid char- 
 acters ; under another set of conditions, what are 
 
THE CLASSIFICATION OF COMPOUNDS 135 
 
 currently accepted as basic characters. For instance, 
 chromium trioxide is usually classed as an acidic oxide, 
 because it reacts readily with the alkalis, partially 
 destroying the characteristic properties of the latter. 
 But it also reacts with sulphuric acid, destroying its 
 characteristic properties, and surely this is a basic 
 function. True, the salt formed in this latter case 
 does not correspond to Cr0 3 , but it has not yet been 
 authoritatively laid down that such correspondence is 
 an essential feature of acids, or bases as the case 
 may be. 
 
 It seems, however, to be tacitly assumed that 
 an oxide has no claim to rank as acidic or basic 
 until it can be proved to produce corresponding salts. 
 Thus, true basic rank was denied to Pb0 2 until the 
 recent preparation of the corresponding salts PbCl 4 , 
 Pb(C 2 H 3 2 ) 4 , &c. In spite of its reaction with hydro- 
 chloric acid, it was regarded as a feeble acidic oxide 
 destitute of true basic properties. Now in virtue of 
 these new salts it ranks as an intermediate oxide. 1 
 
 1 When an acid and base interact to produce a salt only, or a mix- 
 ture of salt and water only, the salt is said to correspond to both the 
 acidic oxide of the acid and the basic oxide of the base. 
 
 BaO + H 2 S0 4 = BaS0 4 + H 2 O 
 
 2 BaO 2 + 2 H 2 SO 4 = 2 BaS0 4 + 2 H 2 O + O 2 . 
 
 According to definition, BaS0 4 corresponds to the oxides S0 3 and 
 BaO, but it does not correspond to the oxide BaOo. If other factors 
 (e.g., oxygen from the air) than the acid and base be involved, the 
 resulting salt is not regarded as corresponding to one or other of the 
 oxides involved. Thus potassium manganate K 2 MnO 4 (regarded as 
 K 2 O MnO 3 ) does not correspond to the oxide MnO 3 from which it is 
 prepared. 
 
i 3 6 LAW AND THEOKY IN CHEMISTRY 
 
 Even if the restriction of correspondence (as denned 
 in the note) were imposed, the difficulties attendant on 
 the classification of oxides would not be greatly lessened. 
 Barium peroxide, for example, though it reacts readily 
 with H 2 S0 4 (S0 3 ) does not form a corresponding salt 
 therewith. Hence it ought not to rank as a basic 
 oxide. But barium peroxide reacts readily with the 
 acidic oxide S0 2 , giving as sole product the salt BaS0 4 ; 
 therefore, from this point of view, the peroxide is a 
 basic oxide. 
 
 When manganese dioxide' is heated with potash in 
 the presence of air, a salt K 2 Mn0 4 is formed. This is 
 said not to correspond to the dioxide Mn0 2 , but to the 
 trioxide MnO 3 , and its formation under these conditions 
 is not regarded as any proof of acidic properties in 
 Mn0 2 . This seems to imply that the Mn0 2 is first 
 oxidised to Mn0 3 by the oxygen of the air, and this 
 acidic oxide then reacts with the potash to form the 
 corresponding salt potassium manganate. But we have 
 no proof that the Mn0 2 does not itself first combine 
 with the potash to form a corresponding manganite, 
 which as soon as it is formed oxidises to manganate. 
 Or, since K 9 is known to be peroxidised when heated 
 in the air, it may be that the exceedingly un- 
 stable salt K 2 Mn0 4 results from the interaction of 
 the acidic oxide Mn0 2 and the basic oxide K 2 2 . 
 But enough has been said to indicate some of the 
 difficulties attendant on the attempt to rigidly classify 
 the oxides. 
 
 Before treating of salts, the subject of neutralisation 
 
THE CLASSIFICATION OF COMPOUNDS 137 
 
 merits a passing notice. When an alkali in aqueous 
 solution is added to an aqueous solution of a strong 
 acid in just the right quantity to destroy completely 
 both the characteristic properties of the acid and the 
 alkali for the two sets of properties disappear simul- 
 taneously the acid is said to be neutralised by the 
 alkali, and vice versd. 
 
 We might employ the disappearance of any one of the 
 characteristic properties of acids or alkalis as an index 
 of the realisation of complete neutralisation, but the 
 property hitherto almost universally selected for this 
 purpose is the reddening action of acids, and the blueing 
 action of alkalis on the vegetable colour litmus the 
 tinctorial matter of a certain species of lichen. The 
 normal hue of this substance is purple, but it turns 
 red when treated with acids, and blue when treated 
 with alkalis. Suppose then we have an acid solution 
 which we wish to exactly neutralise with a solution of 
 alkali, we add a few drops of litmus solution to the 
 acid, and then add the alkali solution cautiously till 
 the red solution containing the acid becomes purple. 
 The neutralisation is then considered to be exact, 
 whereas, if too much alkali had been added, the solu- 
 tion would be blue, if too little, red. Processes of this 
 kind are known generally as processes of titration, for 
 they are usually carried out with, or rather involve 
 somehow, solutions whose strength or titre is known 
 beforehand ; and colouring matters like litmus used as 
 aids to exact titration are called indicators. But the 
 introduction of numerous coal-tar products which give 
 
I 3 8 LAW AND THEORY IN CHEMISTRY 
 
 colour changes with acids and alkalis 1 leads us to ask 
 if the neutral tint of litmus is an index of exact 
 neutralisation, i.e., the presence of acid and alkali in 
 precisely the quantities theoretically necessary for com- 
 plete combination. 
 
 These new artificially prepared indicators are so 
 sensitive that they do not show intermediate neutral 
 tints as is the case with litmus. They not only differ 
 among themselves in sensitiveness to, but also in their 
 qualitative attitudes towards, one and the same solu- 
 tion. Thus, potassium sulphite solution is neutral to 
 phenolphthalein, but changes violet litmus to blue just 
 as an alkali does. Copper sulphate solution is acid 
 as tested by litmus, but neutral in its behaviour to 
 lacmoid. Saliva, which is normally neutral to litmus, is 
 strongly alkaline to lacmoid and acid to turmeric ; and 
 
 1 Among these new indicators may be mentioned 
 
 Phenolphthalein : with acids, colourless ; with alkalis, purple red. 
 Methyl orange : with acids, pink ; with alkalis, pale yellow. 
 Lacmoid : with acids, red ; with alkalis, blue. , 
 
 Rosolic acid : with acids, pale yellow ; with alkalis, violet red. 
 
 The behaviour of a substance in the rfile of indicator, seems to be 
 determined by the balance of acidic and basic characters in its mole- 
 cule. If the acidic properties are relatively strong, the indicator is 
 especially sensitive to bases, and can be successfully used in the 
 titration of salts formed from weak acids, such as carbonates, sul- 
 phides, borates, &c. For carbonic, sulphydric, and boric acids, which 
 are liberated during the process of such titrations, are unable to affect 
 the colour of the indicator. If, on the other hand, basic characters 
 have the predominance in the molecule, then the indicator, being very 
 sensitive to acids, is useless for titrations wherein even such feeble 
 acids as carbonic, sulphydric, &c., are produced. The differences in 
 sensitiveness of these indicators have been applied to the simplification 
 and shortening of many processes of volumetric analysis. 
 
THE CLASSIFICATION OF COMPOUNDS 139 
 
 so on. Indicators, then, differ among themselves in their 
 testifications as to acidity, alkalinity, and neutrality, 
 and there is no reason why the indications of litmus 
 should be accepted in preference to those of other 
 colouring matters. 
 
 Salts. The term salt did not always have its present 
 connotation. It was at one time loosely applied to 
 all substances which tasted like sea-salt, were easily 
 soluble in water and recoverable from their solutions 
 by evaporation. Of these characteristics solubility was 
 regarded as the most important; hence among the 
 alchemists we find an acid referred to as sal acidum, 
 an alkali as sal alkali, while a salt proper was 
 distinguished as sal salsum. It was Eouelle who, 
 totally disregarding solubility relations, appropriated 
 the term salt for the product of the interaction of 
 an acid with a base or metal. The essential feature 
 of an interaction of this kind is the replacement, total 
 or partial, of the replaceable hydrogen of the acid 
 by equivalent quantities of metals. If the whole of 
 the replaceable hydrogen of an acid is displaced by 
 a metal, the resulting salt is in general called a normal 
 salt. Thus Na 2 C0 3 (from the acid H 2 C0 3 Aq) CuS0 4 
 (from H 2 S0 4 ) and Na 3 P0 4 (from H 3 P0 4 ) are all normal 
 salts. The normal salts formed by the interaction of 
 the strong acids (HC1, H 2 S0 4 , HN0 3 , &c.) and the 
 strong bases (KOH, NaOH) were the first to be 
 investigated. As such salts (e.g., K 2 S0 4 , NaCl, KN0 3 , 
 &c.) were all neutral to litmus, it was once customary 
 to regard as synonymous the terms neutral salt and 
 
140 LAW AND THEORY -IN CHEMISTRY 
 
 normal salt. But it is now known that many normal 
 salts are not neutral. When the metal of a weak base, 
 e.g., Fe(OH) 3 replaces all the hydrogen of a strong acid, 
 e.g., HC1, a normal salt with acidic properties results, 
 e.g., FeCl 3 . On the other hand the interaction of a 
 strong base, e.g., KOH and a weak acid, e.g., H 2 C0 3 
 gives a normal salt with alkaline properties, e.g., 
 K 2 C0 3 . 
 
 It' only a. portion of the replaceable hydrogen of the 
 acid is replaced by metal, as in the compounds NaHS0 4 , 
 NaHC0 3 , Na 2 HP0 4 , the resulting salts are called acid 
 salts. This title does not necessarily imply the posses- 
 sion of any other acidic property than the presence of 
 one or more hydrogen atoms replaceable by metals ; as 
 we shall see presently, it does not always imply even 
 this characteristic. It denotes a particular composition 
 rather than a definite set of properties. Thus the 
 slightly alkaline body Na 9 HP0 4 is in terms of our 
 definition an acid phosphate of sodium, while acid 
 sodium carbonate NaHC0 3 possesses very pronounced 
 alkaline properties. 1 
 
 But it is necessary to enlarge our ideas of acid salts. 
 If acid salts were nothing else than acids in which 
 part of the replaceable hydrogen is substituted by an 
 equivalent quantity of metal, then it follows that a 
 dibasic acid like carbonic acid (H 2 C0 3 Aq) should be 
 capable of forming only one acid sodium salt NaHC0 3 . 
 
 1 Acid salts containing replaceable hydrogen and formed from dibasic 
 acids are often distinguished from the normal salts formed from the 
 same acid and base by the prefix bi-. Thus NaHSC>4 is bisulphate of 
 sodium ; NaHCOs bicarbonate of sodium, and so on. 
 
THE CLASSIFICATION OF COMPOUNDS 141 
 
 But there is another way of looking at, and defining, 
 acid salts. In the normal sodium salt of carbonic acid 
 Na 2 C0 3 or Na 2 O.C0 2 
 
 the acid oxide : the basic oxide : : one reacting weight : 
 one reacting weight. 
 
 In the acid salt 2NaHC0 3 or Na 2 0.2C0 2 .H 2 
 
 the acid oxide : the basic oxide : : two reacting weights : 
 one reacting weight. 
 
 In other words an acid carbonate of sodium is a 
 
 i , . , . reacting weights of acidic oxide 
 
 salt 111 which the ratio - 
 
 reacting weights of basic oxide 
 
 exceeds the value unity. This is a wider definition 
 than the one previously given, in that it does not point 
 to any limit for the number of possible acid salts. 
 It prepares us for the statement that a second acid 
 carbonate of sodium exists. This is the so-called 
 sesquicarbonate of sodium which has the composition 
 2Na 2 0.3C0 2 .3H 2 0. 
 
 Every oxy-salt can be conventionally regarded as a 
 compound of basic with acidic oxide. Let us suppose 
 that in the normal salt of a given metal the ratio 
 
 reacting weights of acidic oxide _ 
 reacting weights of basic oxide 
 
 then if other salts formed from the same acid and 
 base or metal exist in which the above ratio has 
 a greater value than n, they must be regarded as 
 acid salts. 
 
142 LAW AND THEOKY IN CHEMISTRY 
 
 We see then that it is not necessary that every acid 
 salt should contain replaceable hydrogen. Thus, in 
 addition to the normal sulphate of antimony Sb 2 (S0 4 ) 3 
 [Sb 2 3 .3S0 3 ], a sulphate having the composition 
 Sb 2 S 4 15 [Sb 2 3 .4S0 3 ] is known. It is therefore 
 in terms of our latest definition an acid salt, and as 
 such it is commonly regarded, though it does not 
 contain replaceable hydrogen. 
 
 Why then, it may be asked, should it not be called 
 a normal salt, seeing that all the hydrogen of the acid 
 from which it has been formed (H 2 S0 4 ) has been re- 
 placed? The answer is, that by tacit consent only 
 those salts are called normal which, in addition to the 
 absence of replaceable hydrogen, can be represented 
 as containing whole multiples of the radicles of the 
 acid forming them. 1 Thus all normal sulphates can be 
 represented as containing electro-positive element united 
 with %(S0 4 ) ; all normal carbonates will contain %(C0 3 ) ; 
 all normal phosphates n (P0 4 ) ; and so on. 
 
 T- ,, ,. reacting weights of acidic oxide 
 
 If the ratio ^ for a 
 
 reacting weights of basic oxide 
 
 particular salt fall below the value n peculiar to the 
 normal salt, then the salt is called a basic salt. In 
 normal mercuric sulphate HgS0 4 [HgO.S0 3 ] n = 1 ; 
 in the sulphate Hg 3 S0 6 [SHgO.SOJ n equals only J. 
 This latter compound is therefore a basic sulphate of 
 mercury. These basic salts constitute quite a large 
 
 1 The molecular formula of acids may be conceived as made up of 
 two parts ; one part consisting of the whole of the replaceable hydrogen, 
 the residue constituting the other part being called the acid radicle. 
 
THE CLASSIFICATION OF COMPOUNDS 143 
 
 and important class of bodies. They are for the most 
 part insoluble substances formed when excess of water 
 acts on the normal salts formed from weak bases, or 
 when excess of weak base is allowed to interact with 
 acid and a strong acid. 
 
 From the fact that there are several acids which 
 do not contain oxygen and have therefore no corre- 
 sponding acidic oxides, it follows that our definitions of 
 acid and basic salts in terms of the ratio 
 
 reacting weights of acidic oxide 
 reacting weights of basic oxide 
 
 lacks generality. We can, however, regard all acid and 
 basic salts from one common standpoint, provided we 
 apply a vaguely extended meaning to the word neutra- 
 lisation. When a weak base interacts with a strong 
 acid, the resulting normal salt as we have seen has 
 in general distinctly acid properties. In virtue of 
 these properties it can interact with more basic oxide, 
 producing what we may conceive of as a nearer 
 approach to absolute neutralisation in the form of a 
 basic salt. According to this view, the weaker bases 
 should show themselves particularly prone to form basic 
 salts with strong acids ; and this is known to be 
 actually the case. 
 
 Again, when a strong base interacts with a weak 
 acid, basic properties predominate in the normal salt, 
 in virtue of which it is capable of further interaction 
 with more acid or acidic oxide, with the realisation of 
 more perfect neutralisation in the form of an acid salt. 
 
144 I AW AND THEORY IN CHEMISTRY 
 
 Thus it follows that acid salts are formed chiefly from 
 the strong bases. 
 
 It has been customary to regard the excess of acid 
 in an acid salt over and above that necessary for the 
 production of a normal salt, as less intimately com- 
 bined or associated with the base than that portion of 
 the acid which just suffices for conditions of normality. 
 Similar remarks apply to excess of base in basic salts. 
 In conformity with these unsubstantiated views, acid 
 and basic salts are often represented as molecular com- 
 pounds of normal salt and acid on the one hand, and of 
 normal salt and base on the other. Thus basic bismuth 
 chloride is sometimes written Bi 2 3 .BiCl 3 [3 BiOCl] ; 
 sesquicarbonate of sodium, according to this method 
 of representation, becomes 2Na 2 C0 3 .H 2 C0 3 .2H 2 0, and 
 acid sulphate of potassium, KgSO^E^SO^ 1 
 
 But if normal salts still retain acid and basic func- 
 tions, enabling them now to combine with more base, 
 now with more acid, the question naturally arises, why 
 should not normal salts with residual acidic functions 
 combine with normal salts possessed of residual basic 
 functions? This question introduces us to the class 
 
 1 According to this view of the matter, the so-called hyperacid salts 
 formed by monobasic acids, e.g., KC1.HC1 or KHC1 2 , KF.HF or 
 KHF 2 &c., ought not to be distinguished from the ordinary acid salt 
 such as Na 2 SO4. HgSO^ The term " hyperacid salts " for these com- 
 pounds is an outgrowth of the primitive views on acid salts, which 
 represented them as acids in which only part of the replaceable hydro- 
 gen is replaced by metals. According to these views it is, of course, 
 impossible for a monobasic acid to form acid salts ; and hence the 
 introduction of the term " hyperacid salts " to meet the necessities of 
 
THE CLASSIFICATION OF COMPOUNDS 145 
 
 of double salts which is at present receiving a good 
 deal of attention. 
 
 Normal sulphate of sodium, Na 2 S0 4 , exhibits re- 
 sidual basic properties in its interaction with sul- 
 phuric acid to produce the acid sulphate of sodium 
 Na 2 S0 4 .H 2 S0 4 (2 NaHS0 4 ). To zinc sulphate, by 
 reason of the existence of such basic salts as 
 S0 3 .2 ZnO, S0 3 .4 ZnO,.2 H 2 0, S0 3 .6 Zn0.10H 2 0, 
 and SOo.8 ZnO. 2 H must be ascribed residual acidic 
 
 o ^ 
 
 properties. What wonder then that basic Na 2 S0 4 
 combines with acidic ZnS0 4 , forming the double sul- 
 phate Na 2 S0 4 .ZnS0 4 .4H 2 ? 
 
 The double halides are an especially important class 
 of bodies, and much is to be hoped from their further 
 investigation. Their formation and existence admit of 
 a provisional explanation similar to that just employed 
 in connection with the double sulphates. 
 
 By reason of the existence of the unstable hyperacid 
 salts having the general formula MX.HX, where X is 
 a halogen and M the metal of an alkali, slight basic 
 properties must be allowed to the alkaline halides. 
 But the normal salts formed by the interaction with the 
 haloid acids of the weaker bases (i.e., the oxides and 
 hydroxides of the heavy metals) have a surplus of acidic 
 characters. Hence we have a large series of com- 
 pounds, called double halides, of the general formula 
 n (MX) m (M'Xa.) where MX represents alkaline halide 
 and M'Xa- the halide of some element other than the 
 metal of an alkali. The following are a few examples, 
 
 picked at random, of such double halides BeCl 2 .2KCl ; 
 
 K 
 
1 4 6 LAW AND THEORY IN CHEMISTRY 
 
 MgF 2 .NaF, PbI 2 .3NH 4 Cl. In accordance with the 
 manner of regarding double salts here developed, it is 
 found that these are most readily formed by halides 
 corresponding to those oxides which in turn most 
 readily form basic salts. Attempts have been made 
 from time to time to remove the so-called double 
 halides from the ill- defined and artificial class of 
 molecular compounds by assigning to them normal 
 unitary atomic structures. 1 These attempts necessi- 
 tated certain extensions of our ideas of valency (see 
 next chapter) which have not met with general 
 acceptance ; and until some more satisfactory theory 
 of the structure of the double halides is framed, these 
 bodies will, in all probability, continue to be empiri- 
 cally regarded as molecular compounds of the simple 
 metallic halides from which they are prepared. 
 
 Inorganic compounds are not all included under the 
 titles acids, bases, and salts. There is in addition a 
 large and rapidly growing number of little-investigated 
 binary bodies, such as the phosphides, the borides, the 
 nitrides, the silicides, the carbides, the selenides, the 
 tellurides, the arsenides, and the antimonides. Most of 
 these substances (which may for the present be classed 
 together under the heading indifferent bodies 2 ) are 
 formed by direct union of the elements at high tem- 
 
 1 For the distinction between molecular and atomic compounds see 
 p. 155. 
 
 2 The oxide NO together with several hydrides, e.g., SiH 4 , PoH 4 , 
 would also fall under this category of indifferent bodies. For an 
 exhaustive list of these indifferent bodies, the reader is referred to 
 Ramsay, System of Inorganic Chemistry, p. 497 ct seq. 
 
THE CLASSIFICATION OF COMPOUNDS 147 
 
 peratures and are decomposed by water. Though in 
 several cases the phosphides, &c., of an element M are 
 of the same type as the phosphides, &c., of the element 
 hydrogen (e.g., P 2 H 6 an( ^ ^V^'s)? J 6 ^ they cann t be 
 regarded as salts of the latter. For the hydrogen 
 compounds do not exhibit the most universal of acid 
 properties the possession of hydrogen atoms directly 
 replaceable by equivalent quantities of other elements. 
 
CHAPTER VI. 
 
 ISOMERISM AND MOLECULAR ARCHITECTURE. 
 
 WITH respect to the arrangement of the simple atoms 
 in the compound atom (molecule) the earlier chemists 
 did not concern themselves. The determination of the 
 mere composition and formula of the compound atom 
 was a problem more than sufficient for their day. But 
 during the years 1820-25 the discovery and investi- 
 gation of cyanic, fulminic, and cyanuric acids forced 
 chemists to face the question of the mutual relations 
 of the atoms in the molecule. These three acids were 
 found to have identically the same percentage composi- 
 tion, and yet the properties of the three bodies differed 
 most pronouncedly. 
 
 So foreign did the possibility of the existence of two 
 or more different bodies of the same percentage com- 
 position appear to the early chemical philosophy, that 
 Berzelius for long refused to admit it. Finally, how- 
 ever, the accumulation of well-authenticated instances 
 of the phenomenon, among which the different tartaric 
 acids may especially be mentioned, not only forced 
 Berzelius to an admission, but led him to introduce the 
 general term "isomerism" 1 for the class of facts 
 under observation. 
 
 1 Much confusion exists in regard to the precise use of the terms 
 
 isomerism, metamerism, &c. Some, regarding identity of percentage 
 
 148 
 
MOLECULAR ARCHITECTURE 149 
 
 What explanation of isomerism could there possibly 
 be other than a difference in atomic arrangements in 
 the molecules of isomeric bodies ? Henceforth it was 
 acknowledged that the properties of a substance depend 
 not only on its composition but also on the atomic 
 architecture of its molecules. Now the architecture of 
 two molecules having identical atomic compositions can 
 differ in two ways. The relative positions of the atoms 
 remaining the same, their distances apart may differ ; 
 or the atomic distances being constant or without in- 
 fluence, the relative positions may differ. It is be- 
 lieved, not without strong evidence, that differences in 
 the relative positions of the atoms constitute the sole 
 determining cause of isomerism ; differences in atomic 
 distances, if indeed such exist, being without influence. 
 
 composition as the sole condition of isomerism, divide isomers into the 
 two classes polymers and metamers ; polymers having different mole- 
 cular weights, metamers the same molecular weight. 
 
 Others again give two meanings a wide and a restricted one to 
 the term isomer. As before, all bodies having the same percentage 
 composition are isomers in the wide sense. Such isomers are then 
 classified into (a) bodies of different molecular weights, (j3) bodies of 
 the same molecular weight. Bodies of the former class are polymers. 
 The latter class is further subdivided into (a') bodies of the same type 
 or isomers in the restricted sense, and (/3') bodies of different types or 
 metamers. 
 
 Throughout this chapter the terms will be used in accordance with 
 Berzelius' first suggestion, viz. : 
 
 Polymers bodies of the same percentage composition and different 
 
 molecular weights. 
 Isomers bodies of the same percentage composition and the same 
 
 molecular weight. 
 Metamers closely related isomers which are capable of very readily 
 
 changing one into another. 
 
 (See article "Isomerism" in Watts' Dictionary of Chemistry.) 
 
i5o LAW AND THEOEY IN CHEMISTRY 
 
 At any rate, all known cases of isomerism can be ex- 
 plained in terms of differences in atomic arrangement 
 alone. 1 
 
 Two, and only two isomers of the formula C 2 H 6 are 
 known ; all attempts to produce a greater number have 
 been futile. But it is obviously possible to arrange the 
 nine atoms in a multitude of different ways. Why then 
 does not a multitude of isomers exist ? There must be 
 a something limiting the possible groupings of a num- 
 ber of atoms in a molecule, whose cause must surely be 
 sought for in the nature of the atoms themselves. 
 
 To take an analogy. Imagine a number of athletes 
 of widely varying strength or holding power. So long 
 as we pay no regard to their holding powers it is 
 evident that we can arrange these athletes in a large 
 number of different groups in each group the relative 
 positions of the athletes being different. 
 
 But the number of possible arrangements is very 
 much diminished as soon as we make the stipulation 
 that the groups are to be conditioned by, and exist in 
 virtue of, the holding powers of the individual athletes 
 groups such as the "human trees" which so fre- 
 quently form one of the items in a circus programme. 
 
 So it is with the atoms in a molecule ; they are not 
 thrown together " higgledy-piggledy " and without any 
 dependence on the intrinsic peculiarities of the atoms, 
 
 1 The distances between those atoms of a molecule which have 
 " strong affinities " for each other may be a determining cause in 
 isomeric change. ButJ the isomcric change from one body A to another 
 body B is a very different thing from the isomerism of A and B. (See 
 p. 183). 
 
MOLECULAK ARCHITECTURE 151 
 
 but the molecule owes its continued existence to the 
 definite holding or linking power of its constituent 
 atoms. To this property or power of atoms the general 
 term valence or valency has been applied. 
 
 As at the birth-time of the recognition of valency, 
 no molecules were known containing per one atom of 
 hydrogen more than one atom of any other element 
 X (i.e., as there were no bodies of the type HX n ), and 
 as the hydrogen atom had already been chosen as the 
 standard for atomic weights, it seemed well to make it 
 also the standard for valencies. Any atom X which 
 held in combination one atom of hydrogen, and formed 
 a molecule HX, was called a monovalent atom. The 
 existence of the molecule H 2 Y established the diva- 
 lency of Y, and so on. 
 
 It will readily be seen that valency is nothing else 
 than a special name given to equivalency when this 
 is applied to atoms, and it is not therefore surprising 
 that the same difficulties as beset the attempt to 
 establish an equivalent system appear again in the 
 later attempts to fix the valencies of the elementary 
 atoms. 
 
 Only a few elements combine with hydrogen to 
 form gasifiable molecules, and had we to rely only on 
 hydrides, our knowledge of the valencies of the atoms 
 of the elements would be very incomplete. Fluorine, 
 chlorine, bromine, and iodine, however, combine atom 
 for atom with hydrogen, hence they are univalent atoms, 
 and may be used as middlemen to fix the valencies of 
 elements which do not form hydrides. Unfortunately 
 
152 LAW AND THEORY IN CHEMISTRY 
 
 the valency of an atom as derived from a study of the 
 halides of the element does not always agree with the 
 valency as fixed by the hydride, e.g., PH 3 , PF 5 . More- 
 over, many elements form at least two gasifiable halides 
 with the same halogen, e.g., HgCl, HgCl 2 . Which 
 particular halide is to determine the valency of the 
 atom in question ? Can we, indeed, correctly speak of 
 the valency of an atom ? l 
 
 This question has given rise to two opposing schools. 
 One school, vaguely referring the phenomena of valency 
 back to some objective attribute of the atom, asserts 
 the necessary constancy of valency. The valency of an 
 atom is, according to this school, as constant as its 
 
 1 From a study of the periodic law, Mendeleeff arrived at the follow- 
 ing generalisation : The sum of the " equivalents " of oxygen and 
 hydrogen with which a single non-metallic atom is combined in its 
 highest salt forming oxide on the one hand, and in its maximum 
 hydride on the other, is constant for all non-metallic atoms, and is 
 equal to 8. Thus, the highest salt-forming oxide of carbon is CO* ; the 
 maximum hydride of carbon is CH 4 . In each case the carbon atom is 
 combined with four "equivalents," and 4 + 4 = 8. The pairs of com- 
 pounds PH 3 , PoO 5 ; AsH 3 , As 2 O 5 ; TeO 3 , TeHo, illustrate the same 
 principle. It is obvious that by " equivalents," Mendeleeff here means 
 something very much akin to valencies ; and in this connection he asks 
 why the valency of an atom should be gauged by its hydrides rather 
 than by its oxides. It should be noted that Mendeleeff' s generalisa- 
 tion is founded on the assumption that peroxides (i.e., oxides of higher 
 types than the so-called group oxides) are not strictly salt-forming do 
 not form corresponding salts. The recent isolation of salts M 2 S 2 O 8 
 (where M monovalent atom) corresponding to sulphur peroxide S-jO? 
 is at variance with this assumption. The sum of the " equivalents " of 
 sulphur in S a O 7 , and SH 2 is 9, and not 8. Some other criterion than 
 the power of forming corresponding salts must lie at the basis of a 
 classification of oxides adapted to Mendeleeff 's generalisation. Further, 
 the partiality of the generalisation, including as it virtually does only 
 the non-metals, is against it. See Mendeleeff 's Principles of Chemistry, 
 vol. ii., appendix 1. 
 
MOLECULAR ARCHITECTURE 153 
 
 weight, although under certain conditions it is not 
 wholly in evidence. Among modern chemists, van't 
 Hoff may be cited as an adherent of this school. He 
 assumes that the attractions in virtue of which the 
 atoms of a molecule hold together, are of the same 
 order as the gravitative attractions of ponderable masses. 
 This being the case, he shows that the intensity of 
 the attraction over the surface of an atom would be 
 constant only if the atom were truly spherical. If, on 
 the contrary, the atom had any other figure than the 
 sphere, then at certain points on its surface the attrac- 
 tive forces would have maximal values these values 
 being unequal inter se. According to van't Hoff's view, 
 then, the valency of an atom expresses the total number 
 of points of maximal attraction this number being 
 dependent on the form of the atom. As the form of 
 the atom is presumably constant, the valency is also 
 constant. If, however, under certain conditions the 
 movements of the atoms conditioning the temperature 
 of the gas become so energetic that only the higher 
 maxima of attraction are powerful enough to come 
 into effective operation, the atom will apparently pos- 
 sess a valency lower than its actual valency. 1 
 
 1 In view of the fact that Faraday's law of electrolysis can be 
 stated in terms of valency data, it may seem somewhat surprising that 
 no satisfactory electrical theory of the cause of valency has appeared. 
 Yet it should be remembered that valency had its origin, and finds 
 its chief application, in organic chemistry, and the great majority of 
 organic compounds, i.e., all organic compounds which are not salts nor 
 acids, are non-electrolytic. In electrolysis the positive electricity may 
 be regarded as conveyed through the electrolyte from anode to kathode 
 by the metallic atoms, the negative electricity from kathode to anode 
 
 UBRAf?y 
 
154 LAW AND THEORY IN CHEMISTRY 
 
 The opposing school, most non-committal and un- 
 imaginative, does not attempt to theorise on, or explain 
 in any way, the phenomena of valency. It does not 
 speak of the absolute valency of any atom, but only 
 of the maximum known valency of an atom or, less 
 general still, of the valency of an atom in a particular 
 compound. 1 
 
 The history of the school of constant valency is an 
 instructive one. Its attempts to vindicate its tenets 
 gave rise not only to such familiar distinctions as those 
 involved in the terms saturated and unsaturated bodies, 
 atomic and molecular compounds, but also to the cele- 
 brated so-called theory of bonds. 
 
 by the non-metallic atoms or acid radicles. Atoms and radicles in 
 their capacity as carriers of electricity during electrolysis are called 
 ions. Faraday's law can be stated in the following form : The elec- 
 trical carrying power of an ion is directly proportional to its valency. 
 In other words, a quantity of electricity, positive or negative, which 
 is conveyed through a salt solution by an ?z-valent atom or radicle, 
 would require for its convection n monovalent atoms or radicles. If a 
 current of electricity be passed through two electrolytic cells in suc- 
 cession, the amount of electricity passing through each cell in a unit 
 of time is the same. If in the first cell we have a salt of a divalent 
 metallic atom, in the second a salt of a monovalent metallic atom, 
 then for every atom (or what is the same thing, for every atomic 
 weight expressed in any mass unit) of the divalent metal deposited in 
 the first cell we shall have two atoms (two atomic weights in the same 
 mass unit) of the monovalent metal deposited in the second cell. See 
 Lodge, Modern Views of Electricity, p. 72 et seq. 
 
 The table given in note 1, p. 159, could quite well be constructed 
 from quantitative electrolytic determinations alone. 
 
 1 Apropos of the attitudes of these rival schools, Ostwald writes : 
 " Fragt man nach dem Wege, auf welchem eine Entscheidung zu treffen 
 wore, so kann eine solche nur auf Grundlage einer bestimmten, wohl 
 begriindeten Hypothese iiber die Natur dessen, was wir Valenz nennen, 
 erlaugt werden." 
 
MOLECULAR ARCHITECTURE 155 
 
 The nitrogen atom combines with three atoms of 
 hydrogen to form NH 3 ; therefore nitrogen is trivaleht. 
 In order to express this fact clearly and succinctly, the 
 notation -/^V was employed. At first the lines proceed- 
 ing from the circle enclosing the symbol for nitrogen 
 were regarded in their true light simply as a species 
 of chemical shorthand ; but soon they came to be 
 regarded as symbolising some objective characteristics 
 of the atoms called bonds or units of affinity. The 
 nitrogen atom was trivalent, because it was possessed 
 of three bonds or units of affinity. It did not hamper 
 the advance of the school, that for long no satis- 
 factory answer could be given to the question, what 
 is the nature of a bond or affinity unit ? But chemists 
 are not the only people who have at times deceived 
 themselves into the belief that to name the unknown 
 is to explain and progress. 1 
 
 But if nitrogen is trivalent, and hydrogen and iodine 
 both monovalent, how, it was asked by the opponents of 
 the school, could the existence of the molecule NH 4 I be 
 explained ? In terms of the difference between atomic 
 and molecular compounds, was the answer. Only in the 
 former class of bodies, it was said, is valency operative. 
 Ammonium iodide is a loose molecular compound of 
 the two atomic compounds NH 3 and HI, in the former 
 of which the nitrogen atom displays its customary 
 
 1 Thanks to van't HofF s theory of valency, we may now, without 
 being accused of a mere glossing of our ignorance, employ the term 
 "bond." Some such term, properly understood, is not only generally 
 useful, but it is absolutely indispensable in describing the results of the 
 new stereo-chemistry (vide infra). The word bond is much to be pre- 
 ferred to its once synonymous term " unit of affinity." 
 
156 LAW AND THEORY IN CHEMISTRY 
 
 trivalency. The attraction (or energy degradation), in 
 virtue of which the atomic compounds HI and NH 3 
 mutually hold each other in combination, is not only 
 independent of, but of a quite different order from, 
 the attractions associated with the valencies of the 
 atoms, and instrumental in holding together the parts 
 of atomic compounds. The former attractions are 
 attributes of the molecules as wholes, and are of a 
 physical rather than a chemical nature. 
 
 However, experiments on the substituted ammonium 
 iodides (NH 4 I in which the H atoms are replaced by 
 different monovalent organic radicles) showed in this 
 special case the artificiality of the distinction between 
 atomic and molecular compounds, and the necessity of 
 regarding ammonium iodide as a true atomic com- 
 pound in which the nitrogen atom, holding in direct 
 combination the four atoms of hydrogen and the one 
 atom of iodine, must be pentavalent. 1 
 
 The school of constant valency was bound to recog- 
 nise the justness of these conclusions, to which it 
 immediately adapted itself by instituting a distinction 
 between saturated and unsaturated compounds, main- 
 taining the while that the nitrogen atom was constantly 
 pentavalent instead of trivalent, as hitherto upheld. 
 
 Here will be observed a slight departure from the 
 original signification of valency ; the valency of an atom 
 is now measured by the number of bonds it possesses. 
 
 1 The behaviours of such presumed inorganic molecular compounds 
 as KT1I 4 [KI.T1I 3 ], K 3 SbBr 3 Cl 3 [SbBr 3 .3K01], &c., are also not in 
 harmony with the deductions which necessarily flow from the admission 
 of molecular combination as distinguished from true atom linking. 
 
MOLECULAR ARCHITECTURE 157 
 
 It at first sight seems more definite than the original 
 signification in that a cause for the phenomena is 
 assigned. But phenomena which are not understood 
 are not cleared up by the mere assigning of causes 
 whose natures are themselves darkly obscure. It is 
 worthy of remark that till van't Hoff's time no satis- 
 factory explanation was advanced of the reason why 
 an atom forms sometimes a saturated, sometimes an 
 unsaturated molecule. 
 
 Conceptions of valency had their origin in the study 
 of carbon compounds, and it is in the field of carbon 
 compounds that the developed theory to-day finds its 
 most successful applications. 1 The reasons for this are 
 not far to seek. The carbon atom is peculiar in that its 
 valency determined from all its highest forms of com- 
 bination containing only a single carbon atom in the 
 molecule, is constant. The molecules CH 4 , CHC1 3 , 
 CHBr 4 , CC1 4 , CS 2 , and C0 2 all point to the tetravalency 
 of carbon. 2 Again, carbon compounds are for the 
 most part easily vaporised, and hence the molecular 
 
 1 It should be particularly noted that organic chemistry, though 
 largely indebted to, is nevertheless quite independent of, the theory 
 of valency. Mendeleeff has shown that the application of Newton's 
 third law of motion in the form in which it appears when regarded as 
 a corollary of the first law, combined with the principle of substitution, 
 is capable of effecting all that the doctrine of valency really effects, viz., 
 the limitation of the number of isomers possible for a given atomic com- 
 plex, and the provision for each isomer of an appropriate molecular 
 ground plan. (See Mendeleeff, Nature, t vol. xl. No. 1032, and Carnegie, 
 American Chemical Journal, vol. xv. No. 1.) 
 
 2 It may be that some day this series will be rendered complete by 
 the discovery of a substance having the molecular composition CN 2 ; 
 the group N 2 being tetravalent, as in the molecule of hydrazine 
 
158 LAW AND THEORY IN CHEMISTRY 
 
 weights of the majority of carbon compounds are 
 known. On the other hand, of only some sixty inorganic 
 compounds of the type MX 7l are the true molecular 
 weights known, and it is impossible to draw unequivocal 
 conclusions regarding the valency of an element M 
 which does not from gasifiable compounds of the type 
 MX n , where X represents some monovalent atom or 
 group. No compound of sodium has been vaporised. 
 Sodium chloride may have the molecular composition 
 NaCl, in which case sodium is a monovalent atom ; but 
 there is no di priori reason why the molecular composi- 
 tion of salt should not be Na 2 01 2 . . . Na ?l Cl n , in which 
 cases the sodium atom may be respectively di- tri- 
 . . . n . . . (2n 1)- valent. 
 
 Further, of the elements with gasifiable compounds 
 of the type MX TO none except silicon, which is very 
 closely related to carbon, shows a constant valency as 
 X is varied. Thus, the phosphorus atom is trivalent 
 in phosphine PH 3 , and pentavalent in phosphorus pen- 
 tafluoride PF 5 . Nor, by reason of the comparative 
 simplicities of the molecules of inorganic bodies, the 
 thorough and so to speak annihilatory nature of the 
 changes they undergo, and the absence of marked 
 cases of isomerism among them, 1 are questions rela- 
 
 1 The only well-marked cases of isomerism (it may, however, be 
 polymerism) in inorganic chemistry are afforded by the so-called 
 inorganic amines the numerous and complex bodies which are formed 
 by the action of ammonia, under varied conditions, on the salts of such 
 metals as platinum, cobalt, chromium, and rhodium. See Ramsay, 
 System of Inorganic Chemistry, p. 524, et seq. The structural formulae 
 which have in some cases been assigned, to these inorganic isomers are 
 open to great doubt and uncertainty. 
 
MOLECULAE ARCHITECTURE 159 
 
 ting to structure of so much importance in inorganic 
 chemistry as they are in the domain of carbon com- 
 pounds. 
 
 For all these reasons the theory of valency admits of 
 much more definite and productive application to carbon 
 compounds than to inorganic bodies; indeed it may 
 seriously be questioned whether the necessarily loose 
 and gratuitous applications of considerations of valency 
 to inorganic compounds have not impeded the progress 
 of the science at large. 1 That such applications brought 
 discredit on the theory itself will be admitted by most. 
 
 1 The valency data of greatest practical importance in inorganic 
 chemistry are here tabulated 
 
 I. Metals and basic radicles 
 
 (a) monovalent, K, Na, Li, Rb, Cs, NH 4 , Ag, Hg ous , Cu ous , 
 
 Au ous , Tl ous , BiO. 
 (j8) divalent, Ca, Sr, Ba, Pb, Mg, Zn, Cd, Hg ic , Cu ic , Sn, 
 
 Fe U8 , Cr ous , Mn ous , Co, Ni, Be, Pt ous , UO 2 . 
 (7) trivalent, Au ic , Al, Bi, Fe ic , Cr ic , Tl ic . 
 (5) tetravalent, Sn ic , U ous , Pt ic . 
 II. Non-metals and acidic radicles 
 
 (a) monovalent - Cl, - Br, - 1, - F, nitrites - NO 2 , nitrates - 
 
 NO 3 , hypochlorites-CIO, chlorates- C1O 3 , cyanides - ON, 
 
 cyanates - CNO, thiocyanates - CNS, metaphosphates - 
 
 PO 3 , perchlorates-C!O 4 . 
 (|8) divalent - O, - S, sulphites - S0 3 , sulphates - SO 4 , thiosul- 
 
 phates- S 2 O 3 , carbonates - CO 3 , silicofluorides-SiF 6 , chro- 
 
 mates - CrO4, dichromates - Cr 2 07, orthomolybdates - 
 
 Mo0 4 , tungstates- WO 4 , metaborates - B 2 O 4 , pyroborates - 
 
 B 4 O 7 , metasilicates - Si0 3 . 
 (7) trivalent, phosphites - PO 3 , phosphates - P0 4 , ferricyanides 
 
 -Fe(CN) 6 , arsenites - AsO 3 , arseniates - AsO 4 , ortho- 
 
 borates - BO 3 . 
 (5) tetravalent, ferrocyanides - Fe(CN) 6 , pyrophosphates - P 2 O 7 , 
 
 orthosilicates - SiO 4 . 
 
 By means of this table, the formula of any common inorganic salt 
 can be found. Place the symbols for the metal and acid radicle of the 
 
160 LAW AND THEORY IN CHEMISTRY 
 
 The chief services that the theory of valency has ren- 
 dered the chemistry of carbon compounds have been in 
 the domain of isomerism. The theory enables us to pre- 
 dict with absolute certainty the number of isomers con- 
 sistent with a given molecular composition. The general 
 idea one is apt to glean from much of the literature 
 on the subject is that valency lies at the basis of the 
 so-called constitutional or structural formulae. This is 
 not the case. What does lie at the basis of structural 
 formulas is merely the conception of definite atomic 
 linkage a conception of lower order than that of 
 valency. It is quite possible to symbolise succinctly the 
 chemical relations of a body to give it a structural 
 
 salt in juxtaposition ; then multiply one or other of these symbols, or 
 it may be both, so that the metallic atoms and acid radicles become 
 equivalent. Express these multiples in small subscripts placed after 
 the respective symbols, and the result is the formula required. Thus 
 calcium phosphate is Ca 3 (PO^o ; for we must take three divalent atoms 
 and two trivalent radicles in order to get a balance of the valencies of 
 the two constituent parts of the salt. 
 
 The table is necessarily incomplete. Elements which form halides 
 of several types, but no stable oxysalts, e.g., W and Mo, have been 
 omitted. As regards the formulae of oxides, the table is only available 
 for those markedly basic oxides which form corresponding salts ; it 
 correctly gives the formulae for such oxides as ferrous oxide, baric 
 oxide, &c., but it does not help us in the cases of the magnetic oxide 
 of iron, baric peroxide, &c. Further, it only gives the formulae for 
 normal salts ; the complex poly-salts formed by weak acids such as 
 boric, molybdic, tungstic, silicic, &c., are most simply represented, not 
 as made up of metal and acid radicle, but of acidic and basic oxide, 
 thus n BO m AO, where BO is the oxide of a metal, AO the oxide of 
 a non-metal, and m n are whole numbers. However, these complex 
 salts can also be represented as derived from the normal salts. Thus 
 the mineral serpentine usually written 2SiO.>3MgO, may also be 
 regarded as a dehydrated acid magnesium ortho-silicate, Mg 3 H 2 
 (SiO4) 2 -H. 2 O = Mg 3 Si:>07. Obviously, the table does not inform us as 
 to the possible number or the limit forms of such complex salts. 
 
' MOLECULAR ARCHITECTURE 161 
 
 formula recalling its mode of formation, suggesting its 
 decompositions, &c., by means of certain conventions 
 respecting the written arrangements and groupings of 
 the constituent atoms, and without in any way having 
 recourse to the teachings of valency. Thus, the symbol 
 
 
 HO . S . OH conveys much information concerning sul- 
 
 
 
 phuric acid, and yet is framed without any regard to 
 the doctrine of valency. This symbol pictures the fact 
 of the symmetry of the sulphuric acid molecule, and 
 therefore the impossibility of the existence of isomeric 
 forms of the derivatives HBS0 4 and RR'S0 4 . 
 
 Further, knowing the equal value, substitutionally, 
 of Cl and (OH), of H and K, &c., and also the readiness 
 with which hydroxyl groups linked to the same atom 
 interact, the following transformations are at once 
 suggested by the symbol given 
 
 00 00 
 
 C1.S.C1 Cl.S.OH Cl.S.OK KO.S.OK S.O 
 
 0000 
 
 To argue whether the sulphur atom is tetra- or hexa- 
 valent in sulphuric acid is profitless. The formulas 
 HO .0 
 
 HO S OH S 
 
 HO 
 
 do not convey any more information about the pro- 
 perties of sulphuric acid than does the simple 
 
 
 
 representation HO . S . OH. 
 O 
 
1 62 LAW AND THEORY IN CHEMISTRY 
 
 Before proceeding to illustrate the applications of 
 considerations of valency to organic bodies, it is first 
 necessary to speak of the valencies of compound 
 radicles. A compound radicle is a group of atoms 
 which, like a single elementary atom (simple radicle) 
 appears as an undecomposed whole throughout a large 
 series of compounds or transformations. Like an atom, 
 too, a compound radicle has a definite valency. These 
 compound radicles, or stable atom complexesj play so 
 conspicuous and important a role in the field of carbon 
 compounds that they were once regarded as the true 
 atoms of organic chemistry, which itself was supposed 
 to be adequately defined as the chemistry of compound 
 radicles. As random examples of these compound 
 radicles we may cite ethyl C 2 H 5 , ethyoxyl 2 H 5 0, 
 hydroxyl OH, carboxyl COOH, and amidogen NH 2 
 all monovalent radicles, for each combines with a 
 single atom of hydrogen to form a gaseous mole- 
 cule ; the divalent radicles ethenyl C 2 H 4 , carbonyl 
 CO, and the trivalent radicles, glyceryl C 3 H 5 and 
 methenyl CH. 
 
 Now, if the carbon atom can hold in combination 
 four hydrogen atoms, it will also hold in combination 
 four monovalent compound radicles, and therefore such 
 molecules as 
 
 OC 2 H 5 H NH 2 
 
 I I I 
 
 C 2 H 5 - C - OCoH 5 HOOC - C - COOH H - C - COOH, &c. 
 I I I 
 
 OC 2 H 5 H H 
 
 ought to, and do, exist. 
 
MOLECULAR ARCHITECTURE 163 
 
 Hence we are bound to give an extended significance 
 to the statement the carbon atom is tetravalent. Not 
 only does it imply that the carbon atom combines with 
 four atoms of hydrogen to form a gaseous molecule, 
 but further that in any molecule a carbon atom is 
 capable of directly interacting with four, but no 
 more than four, other atoms, the particular valencies 
 of these atoms being a matter of indifference. 
 Thus in the ethyl ether of ortho-carbonic acid, 
 C(OC 2 H 5 ) 4 , the heavily printed carbon atom directly 
 interacts with four divalent oxygen atoms, each of 
 which again directly interacts with a monovalent 
 ethyl group. In glycocine 
 
 CH 2 .KH 2 .COOH 
 
 the heavily printed carbon atom acts on two mono- 
 valent hydrogen atoms, one trivalent nitrogen atom, 
 and one tetravalent carbon atom four atoms in all. 
 
 Let us now proceed to apply these considerations to 
 isomerism. A body is discovered having the molecular 
 composition C 2 H 6 0. Can it have isomers, and if so, 
 how many? The answer returned by the valency 
 theory is this. If the constituent atoms can be arranged 
 in n different ways so that each of the carbon atoms 
 never interacts directly with more than four other 
 atoms, and each of the oxygen atoms with more 
 than two other atoms, then n isomers, all told, are 
 possible. As a matter of fact, it is found on trial 
 that the atoms C 2 H 6 can be arranged in only 
 
164 LAW AND THEOKY IN CHEMISTRY 
 
 two ways, subject to the restrictive conditions just 
 stated 
 
 H H H H 
 
 I I II 
 
 H C C H H C C H 
 
 II II 
 H H H H 
 
 In strict accordance with this result, all attempts to 
 prepare more than two isomers of the composition 
 C 2 H 6 have failed. The two known substances having 
 this formula are methyl ether and ordinary alcohol. 
 
 The problem next rises, which of these formulse are 
 we to ascribe to methyl ether ; which to alcohol ? This 
 is solved by the knowledge of the fact that one atom 
 of hydrogen in the alcohol molecule differs from the 
 remaining five in that it alone can be replaced by 
 alkali metals. Now, in the left-hand formula all the 
 atoms of hydrogen are similarly related to the molecule, 
 but in the right-hand formula five atoms of hydrogen 
 directly interact with carbon atoms (i.e., are directly 
 linked to carbon atoms), while the sixth hydrogen atom 
 acts only indirectly on a carbon atom through the 
 oxygen atom. Therefore the right-hand formula belongs 
 to the molecule of alcohol, and therefore by exclusion 
 the left-hand one represents methyl ether. Further, 
 the syntheses and all the known transformations of 
 methyl ether, and ethyl alcohol, are in strict conformity 
 with, and indeed are deducible from, their appropriate 
 symbols thus arrived at. 
 
 Although in many cases more isomers ought to exist 
 
MOLECULAR ARCHITECTURE 165 
 
 according to the teachings of valency than are at 
 present known, 1 yet in no case is a greater number of 
 isomers known than is provided for, and anticipated 
 by, the valency theory. 
 
 Till within quite recent years this statement could 
 not have been made. 
 
 It would seem to follow from the simple case already 
 considered, that isomerism ought always to be accom- 
 panied by differences in chemical properties, these 
 differences finding expression in, and being due to, 
 the differing molecular architectures of the isomers. 
 Yet bodies are known which, having the same molecular 
 weight and the same composition, undergo precisely 
 the same chemical transformations, and are yet quite 
 different in such physical properties as crystalline form, 
 solubility, action on polarised light, &c. Such bodies, 
 on account of the identity of their chemical transfor- 
 mations, must be assigned identical formulae identical 
 molecular architectures ; and for this reason they were 
 for long regarded as forming exceptions to, or at least 
 falling outside the pale of the province of the "theory" 
 of valency. Hence they were relegated to a kind of 
 suspense class, bearing the title physical isomers. 
 
 We now know, thanks to the researches of van't 
 Hoff and Le Bel, that these are not exceptional pheno- 
 mena, but are fully provided for by the theory of 
 valency, if we do not limit the possible different 
 
 1 For instance, nine heptanes having the composition CrHie, ought 
 to exist according to the teachings of valency ; as yet, however, only 
 four bodies having this composition have been isolated. 
 
1 66 LAW AND THEORY IN CHEMISTRY 
 
 arrangements of atoms in space by the different arrange- 
 ments possible in a plane. 
 
 It is not irrational to assert that two isomers t of the 
 formula CH 2 C1 2 should exist. For the formulae 
 
 H H 
 
 I I 
 
 Cl C Cl Cl C H 
 
 I I 
 
 H Cl 
 
 while both in accordance with the dictates of valency, 
 obviously differ. In the right-hand formula the two 
 chlorine atoms are not related to the rest of the mole- 
 cule in the same way as they are in the left-hand 
 formula. Why should not this difference entail a 
 difference in properties a possible isomerism of the 
 molecule CH 2 C1 2 ? It evidently does not, for all 
 attempts to prepare two different bodies of the formula 
 CH 2 C1 2 (and Henry has made many such) have failed. 
 We are therefore compelled to still further enlarge 
 our conceptions of valency by the conclusion that there 
 is really no difference between the above formulas that 
 the order of cyclical arrangement of the atoms round 
 the central carbon atom is, in effect, immaterial. Con- 
 sequently, no isomeric molecules of the types CHR r 3 , 
 CHR T R n 2 , CHR'^'R 111 , CR^R'V CR^R 11 , CR 1 R ir R III R IV , 
 should exist. In partial accordance with this con- 
 clusion, no isomers of the first five types are known, 
 but, contrary to the conclusion, all the simplest of the 
 so-called physical isomers conform to the sixth type, 
 
MOLECULAR ARCHITECTURE 
 
 167 
 
 " 1 !^ possess what is called an asymmetric 
 carbon atom. 
 
 Van't Hoff and Le Bel showed the way out of this 
 difficulty by reminding us that molecules are tri- and 
 not di-dimensional entities. 
 
 We get a spatial conception of molecules in accord- 
 ance with the limitations of valency, if we conceive the 
 carbon atom placed at the centre of a tetrahedron, its 
 four "valencies" or "bonds" being directed towards 
 the four solid angles a, b, c, 
 d, of the tetrahedron. 
 
 So long as the monovalent 
 atoms or radicles on which 
 the carbon atom directly in- 
 teracts are not all different, 
 then it is impossible to 
 place them at the angles a, 
 I, c, d, so as to occupy in dif- 
 ferent dispositions different 
 
 spatial relations to the central carbon atom. Let A 
 stand for one of the dispositions, and B for another, 
 and suppose the two tetrahedra interpenetrative, then 
 it will always be possible, no matter the dispositions, so 
 to place A inside B, or vice versd, that the similar atoms 
 or radicles attached to the central carbon atoms will 
 fall together and occupy the same region of space. 
 In short, the two dispositions will always be super- 
 posable. 
 
 But when the atoms or groups at the angles a, b, c, d, 
 are all different from one another, then it is possible 
 
 F IG . 3. 
 
1 68 LAW AND THEORY IN CHEMISTRY 
 
 to get two dispositions which are non-superposable 
 the one disposition bearing to the other the same rela- 
 tion as does an object to its image in a plane mirror ; 
 the same relation as a left-hand glove bears to a right- 
 hand one. In crystallography this would be called an 
 enantiomorphous relationship. In other words, if in 
 one molecule the atoms or radicles are regarded as 
 disposed around the central carbon atom in a right- 
 handed spiral, then in the other molecule the disposition 
 is a left-handed spiral. 
 
 Let us illustrate this last case by the lactic acids 
 which indeed first suggested to Wislecenus the neces- 
 sity of introducing stereometric conceptions into the 
 domain of molecular architecture. 
 
 Three lactic acids are known ; they have all practi- 
 cally identical chemical properties, but differ from each 
 other in their action on polarised light, and in the 
 solubilities of their salts. Dextrolactic acid turns the 
 plane of polarisation to the right; sarcolactic acid to 
 the left; while fermentation lactic acid is optically 
 inactive. The molecules of these lactic acids must, from 
 their chemical relationships, have the structure 
 
 CH 3 
 
 HOOC C H 
 
 I 
 OH 
 
 Th,e heavily printed carbon atom in the above symbol 
 is obviously asymmetric. Therefore there are possible 
 
MOLECULAR ARCHITECTURE 
 
 169 
 
 two different dispositions around it of the radicles it 
 holds in combination, viz. : 
 
 OH 
 
 COOH 
 
 COOH 
 
 OH 
 
 FIG. 4. 
 
 FIG. 5. 
 
 One of these dispositions must characterise dextrolactic 
 acid ; the other, sarcolactic acid. Yet in the present 
 state of our knowledge it is impossible to definitely 
 allocate them. 
 
 But what of the third lactic acid the inactive 
 modification ? The theory at first sight does not seem 
 to provide for it. 
 
 It has now been definitely proved that inactive or 
 ordinary lactic acid is not a true chemical unit, but 
 a mixture in compensating quantities of the dextro 
 and laevo modifications. Its formula would be given 
 by placing a plus sign between the above two figures. 1 
 
 Attempts have recently been made to find some con- 
 
 1 For the further development of stereometric ideas we would refer 
 the reader to Marsh, Chemistry in Space, and Auwers, Die Entivickelung 
 der Stereochemie. 
 
 Suffice it here to say that there are no known instances of physical 
 isomersion of carbon compounds which do not receive full and adequate 
 interpretation in terms of the attributes of the asymmetric carbon atom. 
 
i;o LAW AND THEOKY IN CHEMISTRY 
 
 nection between the masses of the radicles, R 1 , E/ 1 , R m , 
 and R IV , held in combination by an asymmetric carbon 
 atom, and the nature and amount of the optical activity 
 which, unless compensatory influences come into play, 
 is invariably associated with asymmetric carbon atoms. 
 A regular tetrahedron has six planes of symmetry ; 
 a plane of symmetry bisecting each of the six inter- 
 facial angles of the figure. 1 Let us suppose that the 
 centre of such a tetrahedron is occupied by a carbon 
 atom, and that four atoms or radicles are disposed round 
 the carbon atom so that their masses are concentrated 
 at the apices of the four solid angles of the tetrahedron. 
 So long as these four radicles are not all dissimilar, the 
 centre of gravity of the whole system will be in one 
 or other of the six planes of symmetry of the tetra- 
 hedron, and there will be no optical activity. But 
 when all the radicles or atoms differ among themselves, 
 the centre of gravity of the system no longer falls 
 within any of the planes of symmetry, and optical 
 activity makes its appearance. Let the perpendicular 
 distances of the centre of gravity of the system in 
 this latter case from the six planes of symmetry be 
 d v d 2 , d 3 , dy d & and d 6 . Then the product P = d l X d 2 
 X d 3 x d X d 5 x d 6 is, according to Guye, a measure of 
 
 1 " A plane of symmetry may be defined as a plane which is capable 
 of dividing a body into two halves which are related to each other in 
 the same way that an object is to its reflection in a mirror. More 
 exactly we may say : two objects or two halves of the same object are 
 symmetrical with reference to a plane placed between them, when from 
 any point of one object a normal to this plane, prolonged by its own 
 length on the opposite side of the plane, will meet the corresponding 
 point of the other object." Williams, Elements of Crystallography. 
 
MOLECULAR ARCHITECTURE 171 
 
 the asymmetry of the molecule, and should therefore 
 be proportional to the optical activity of the substance. 
 If plus and minus signs be conventionally applied to the 
 distances d v d 2 , d 3 , &c., according as they are measured 
 from one side or the other of each plane of symmetry, 
 P may be either positive or negative according to 
 circumstances. 
 
 Guye has shown how P may be calculated before- 
 hand for any meditated derivative of a given optically 
 active substance. If the passage from an optically 
 active body A to an optically active substitution 
 derivative A' would involve an increase in the value 
 of P, then, according to Guye, we may predict that 
 A' will be more active optically than the parent 
 substance ; if the transformation would involve a 
 diminution in the value of P, then the new body 
 will be less active than the original one from which 
 it was derived. Finally, if the passage from A to 
 A' is accompanied by a change of sign of P, then 
 the parent substance and its derivatives* will produce 
 opposite rotations; one will be a dextro-rotatory sub- 
 stance, the other a laevo-rotatory substance. 
 
 In the following symbol for active amyl chloride, 
 
 it can readily be shown that CH CH 
 
 the centre of gravity of the 
 system falls on the CH 2 C1 \\ 7 c / a cl 
 
 side of the dotted plane of 
 symmetry C 2 H 5 mH. If there- 
 fore the chlorine atom in the 
 CH 2 C1 group be replaced by a heavier atom, the 
 
172 
 
 LAW AND THEORY IN CHEMISTRY 
 
 HO 
 
 centre of gravity of the system must still remain on 
 the same side of the plane C 2 H 5 mH, but will be 
 further removed from it. Accordingly it is found 
 that amyl bromide has higher optical activity than 
 the chloride ; and amyl iodide in turn higher activity 
 than the bromide. 1 
 
 Let us take another illustrative application of Guye's 
 theory. 
 
 In the accompanying symbol for dextro-tartaric acid, 
 the centre of gravity would 
 lie on the C0 2 H side of the 
 dotted plane Hmm'H. Let 
 us now replace the H atoms 
 of the hydroxyl groups by 
 acetyl radicles (acetyl = 
 CH 3 CO). This substitution 
 carries the centre of gravity 
 over to the other side of the 
 plane Hmm'H ; and it is 
 found as a matter of fact 
 that diacetyl tartaric acid is 
 laevo-rotatory. 
 Now let us form etherial salts of diacetyl tartaric 
 
 1 The values are as follows : 
 
 [o] D = 1 6' for the chloride. 
 = 4 24' for the bromide. 
 = 8 20' for the iodide. 
 
 In these equations [ct] D may be taken to represent the angle through 
 which the plane of a polarised ray of sodium light would be turned if 
 it were made to pass through a tube of y^ square centimetre cross 
 section, and of just sufficient length to hold exactly one gram of the 
 substance under examination. 
 
 FIG. 
 
MOLECULAR ARCHITECTURE 
 
 173 
 
 acid by replacing the H atoms of the carboxyl groups 
 with alkyl monovalent radicles E (e.g., C 2 H 5 , C 3 H 7 , &c.). 
 This brings the centre of gravity of the system back 
 towards the C0 2 E side of the plane Hmw'H, and 
 consequently we find that the Isevo-rotatory power 
 of the etherial salt diminishes as the weight of R 
 increases, until finally the rotation changes sign and 
 the higher etherial salts of diacetyl tartaric acid are 
 dextro-rotatory. 
 
 Diacetyltartaric acid [a] D = - 23'14 
 
 Methyl-diacetyl-tartarate - 14-29 
 
 Ethyl + 1-02 
 
 Propyl + 6-52 
 
 Although many facts are in harmony with pre- 
 dictions founded on Guye's views of the relations 
 between optical activity and molecular asymmetry, 
 
 CH 
 
 CH OH 
 
 HO 
 
 COOC 
 
 iP, 
 
 OH 
 
 FIG. 8. 
 
 FIG. 9. 
 
 yet isolated instances of exceptions to his generalisa- 
 tions are not wanting. Thus from the accompanying 
 diagrams it is easy to see that ethyl gly cerate (Fig. 9) 
 
174 LAW AND THEORY IN CHEMISTRY 
 
 should have higher optical activity than ethyl lactate 
 (Fig. 8) ; but as a matter of fact the optical constants 
 for these two bodies are as follows : 
 
 Ethyl lactate [a] D = - 14-19 
 Ethyl glycerate [a] D = - 9-18 
 
 Again, such a body as ethylic diacetyl glycerate 
 
 CH 2 O.C 2 H 3 
 H.C.O.C 2 H 3 
 C OOC 2 H 5 
 
 should be optically inactive if mass were the only factor 
 determining optical asymmetry ; for the two isomeric 
 radicles CH 2 . . C 2 H 3 and COOC 2 H 5 have equal 
 masses, and when this is the case P must be zero. For 
 if a, I, c, d represent the masses of the radicles held in 
 combination by an asymmetrical carbon atom in such 
 a way that these masses are concentrated at the apices 
 of a regular tetrahedron, then one of the factors in 
 determining the value of P is 
 
 (a - b)(a - c) (a - d) (b - c) (b - d) (c - d) 
 (a + b + c + d)* 
 
 It is clear that if any two of the masses a, b, c, d 
 become equal in value, then P the product of 
 asymmetry must also become zero. When this is the 
 case, the optical activity which is supposed to accom- 
 pany asymmetry in Guye's sense, should disappear. 
 Hence it would appear that in addition to mere mass 
 
MOLECULAR ARCHITECTURE 175 
 
 relationships, considerations of structure must also find 
 a place in future attempts to quantitatively link to- 
 gether optical activity and molecular asymmetry. 
 
 So far we have only dealt with what were once 
 regarded as exceptional isomerisms among saturated 
 carbon compounds. We now turn to peculiar cases of 
 isomerism presented by unsaturated carbon compounds 
 of ethylenic and acetylenic types, i.e., compounds in 
 whose molecules one or more of the carbon atoms 
 directly acts on less than four other atoms or radicles. 1 
 Many of the isomeric bodies of this class, while on 
 the whole resembling each other chemically, yet in addi- 
 tion to purely physical differences manifest also minor 
 chemical contrasts. It is of especial significance that 
 the physical differences alluded to do not, as in the 
 case of saturated bodies, involve optical activity. ' 
 
 1 The carbon atoms in ethylene and acetylene are often said to be 
 doubly and trebly linked respectively, and the two substances are 
 represented as follows 
 
 H-C-H C-H 
 
 II ' 111 
 
 H-C-H C-H 
 
 C 
 If | symbolises a certain interaction between a pair of carbon 
 
 C 
 
 C C 
 
 atoms, i| and ||| certainly suggest interactions of double and treble 
 
 C C 
 
 intensity ; but Thomsen's thermochemical studies in this direction, as 
 well as the salient chemical characteristics of ethylene and acetylene, 
 are at variance with such a state of things. That all the four points of 
 maximal attraction ("bonds") of the carbon atom must always be in 
 active operation is an unwarranted assumption passed down from the 
 early school of constant valency. The intra-molecular movements of 
 ethylene may be of such a nature that only the three greatest maxima 
 of attraction are able to effectually assert themselves. 
 
1 76 LAW AND THEORY IN CHEMISTRY 
 
 Moreover, these cases of isomerism, unlike the optical 
 isomerism just discussed, are not inconsistent with the 
 two-dimensional expression of the doctrine of valency. 
 As illustrative, let us take the case of the isomerism of 
 maleic and fumaric acids. 
 
 These two acids differ in their physical properties as 
 follows. While fumaric acid sublimes on being heated, 
 maleic acid has a definite melting point, 130 ; and the 
 latter acid is much more soluble than the former. 
 
 While the chief reactions 1 of the two acids demand 
 in each case the same rational formula, viz. 
 
 fC.H.COOH 
 I C.H.COOH 
 
 yet the following minor differences between the two 
 acids may be cited. Fumaric acid is more stable than 
 maleic, so that reactions which take place with maleic 
 acid, under ordinary conditions, require high tempera- 
 tures and high pressures in the case of fumaric acid. 
 Some reactions, e.g., etherification, proceed in the case 
 of both acids under the same conditions, but at a slower 
 rate with fumaric than with maleic acid, Maleic acid 
 readily yields an anhydride, fumaric acid does not. The 
 action of bromine on the two acids gives rise to different 
 products; from fumaric acid dibromosuccinic acid re- 
 sults ; from maleic, iso-dibromosuccinic acid. Another 
 peculiarity of these isomers is the readiness with which 
 fumaric acid can be changed into maleic acid derivatives, 
 
 1 Both acids are dibasic, and are formed from malic acid by dehydra- 
 tion. Reducing agents convert both acids into succinic acid. 
 
MOLECULAR ARCHITECTURE 177 
 
 and maleic acid into fumaric acid derivatives. Thus, 
 if maleic acid is treated with bromine, and then the 
 elements of hydrobromic acid are subsequently removed 
 by the action of water, the result is bromofumaric acid. 
 Conversely, bromomaleic acid can be obtained by a 
 similar series of operations from fumaric acid. 
 
 Enough has been said to indicate that we have here 
 a much more pronounced kind of isomerism than that 
 which the lactic acids presented. Moreover, in this 
 case different di-dimensional arrangements of the 
 atoms in the complex C 2 H 2 (COOH) 2 are consistent 
 with the tetravalency of the carbon atom. Hence 
 numerous attempts have been made to assign to the 
 two acids di-dimensional structural formulae. Such 
 structural formulae necessarily involve different radicles 
 in the two cases, despite the great chemical similarity 
 of the acids. Thus, Anschiitz, while retaining the 
 formulae 
 
 COOH C H 
 
 II 
 COOH C H 
 
 for fumaric acid, advocates 
 
 as best representing the molecular, structure of maleic 
 
 acid. But these di-dimensional representations met 
 
 M 
 
iy8 LAW AND THEORY IN CHEMISTRY 
 
 with so little favour generally that the term allo 
 isomerism was provisionally introduced to group to- 
 gether such apparently inexplicable cases of isomerism 
 among unsaturated bodies as are typified by the acids 
 under discussion. 1 
 
 When recourse is had to spatial considerations we 
 find that all difficulty disappears, and that the reten- 
 tion of the term allo-isomerism is needless. For there 
 are two, and only two, tri-diniensional arrangements 
 
 COOH 
 
 HOOC 
 
 COOH 
 
 FIG. 10. FIG. 11. 
 
 of the atom complex C 2 H 2 (COOH) 2 , in each of which 
 essentially the same structural units obtain, i.e., the 
 same radicles are involved in each case. These two 
 arrangements will be sufficiently obvious from the 
 accompanying plane projections. 
 
 So far our conceptions of valency have been purely 
 
 1 There are, however, many who still maintain that the di-dimen- 
 sional formulae represent the properties and peculiarities of the two 
 acids better than do the accepted tri-dimensional ones. 
 
MOLECULAR ARCHITECTURE 179 
 
 statical, but Wislicenus, by introducing dynamical 
 ideas, has much extended the original theory of van't 
 Hoff and Le Bel. According to Wislicenus, the atoms 
 in a molecule exert influences on each other even when, 
 according to the teachings of valency, they do not 
 directly interact, i.e., link each other. The negative 
 or chlorous atoms have a great attraction or affinity 
 for the positive or basylous atoms, and this attrac- 
 tion is either partially satisfied in the molecule 
 by the chlorous atoms swinging themselves as near 
 as possible to the basylous atoms ; or it may be 
 that the attraction only sets up an intra-molecular 
 stress which, under favourable conditions of tempera- 
 ture, &c., asserts itself, and causes such a rotational 
 movement in the molecule that the mean distance 
 between the mutually attracting atoms is made as 
 small as possible. 
 
 In terms of these views, not only is it possible in 
 many cases to assign the appropriate formula to a 
 given isomer, but obscure chemical transformations, 
 such as the before-mentioned mutual convertibility 
 of fumaric and maleic acid derivatives, find a full 
 explanation. Indeed Wislicenus' views have raised 
 the van't Hoff-Le Bell hypothesis from the level of 
 mere co-ordination and explanation of known facts, 
 to the higher level of prophecy; it now not only 
 explains, but anticipates facts. For a complete exposi- 
 tion of Wislicenus' views we would refer the reader 
 to the pamphlet Uber die rdumliche Anordnung der 
 Atome in organischen Molekulen (Hirzel, Leipzig). 
 
 - UNIVEBSITl 
 
i8o LAW AND THEORY IN CHEMISTRY 
 
 All we can do here is to exemplify these new views 
 very briefly as they bear on the isomerism of maleic 
 and fumaric acids. 
 
 The first question to be settled is Of the two 
 formulae, which is to be assigned to fumaric, which 
 
 f -cooj 
 
 HO O C 
 
 to maleic acid ? The left-hand formula figures a 
 more stable system than does the right-hand one, 
 for in the former the chlorous carboxyl groups are 
 as near as is possible to the basylous hydrogen atoms. 
 But fumaric acid is, as we have said, a more stable 
 acid than maleic; hence the left-hand formula sym- 
 bolises fumaric acid, and, by exclusion, the right-hand 
 formula maleic acid. The constitution of maleic acid 
 explains its ready dehydration. The water which 
 splits off during the formation of an anhydride from 
 an acid is known to result from the hydroxyl portions 
 of carboxyl groups, and it is but natural to assume 
 
MOLECULAR ARCHITECTURE 
 
 181 
 
 that the proximity in the molecule of these groups 
 would favour a reaction in which both are simul- 
 taneously implicated. 
 
 Let us now explain by a typical example the rationale 
 of the formation of maleic acid derivatives from f umaric 
 acid. When bromine and fumaric acid are heated to- 
 gether, each molecule of the acid takes up two atoms 
 of halogen. This, in terms of stereometric formulae, can 
 only take place in one way : 
 
 HOOC 
 
 HOOC 
 
 COOH 
 
 COOH 
 
 BJ? 
 
 FIG. 14. 
 
 But the resulting molecule pictured is in a state of 
 internal stress by reason of the attractions of the 
 strongly positive hydrogen atoms for the strongly 
 negative bromine atoms. As a result of this stress, 
 an intra-molecular rotation around the axis joining the 
 two asymmetric carbon atoms ensues, so as to bring 
 the hydrogen atoms into as close proximity as possible 
 
1 82 
 
 LAW AND THEORY IN CHEMISTRY 
 
 to the bromine atoms. 1 Hence, when by the subse- 
 quent action of water a molecule of hydrobromic acid 
 
 COOII 
 
 HBi* 
 
 COOH Br 
 
 FIG. 15. 
 
 COOI 
 
 COOH 
 
 is removed, the only possible result is bromo-maleic 
 acid. 
 
 1 From the phenomena presented by the isomerism of the three 
 
 HON.C.C 6 H 5 
 benzil dioximes, all of which have the formula Meyer 
 
 HON.C.C 6 Hg 
 
 and Auwers concluded that carbon atoms may be linked in two ways ; 
 one way admitting of free rotation of the parts of the molecule round 
 the joining axis, and another way inconsistent with any rotational 
 movement. It has, however, been pointed out, that the exceptional 
 isomerisms exhibited by the benzil dioximes might be explained 
 without having recourse to the idea of non-rotational, singly linked, 
 carbon atoms, by applying stereometric considerations to the nitrogen 
 atom e.g., regarding it as occupying one angle of a tetrahedron, its 
 three " valencies " or "bonds" being directed towards the other three 
 angles. But as yet nothing very definite can be said concerning the 
 stereochemistry of the nitrogen atom ; the whole subject has not yet 
 emerged from the purely tentative stage. It may here be stated that 
 an attempt has quite recently been made to account for the isomerism 
 exhibited by amido-platinum compounds of the type 
 
MOLECULAR ARCHITECTURE 183 
 
 Many substances are known, to each of which it 
 seems necessary to assign more than one formula. 
 Such substances therefore exhibit a peculiar kind of 
 isomerism which has been called tautomerism. Some 
 of the reactions of such a tautomeric body suggest 
 one structure for it, while other of its reactions seem 
 to demand quite another structure. It has been pro- 
 posed to call the alternative structures of a tautomeric 
 body its desmotropic forms or states. 
 
 Laar, who first drew attention to this form of 
 isomerism, is inclined to think that the molecular 
 architecture of a tautomeric body is changing from 
 moment to moment that the structure of hydrocyanic 
 acid is at one instant 
 
 HNC 
 and at the next instant 
 
 NCH. 
 
 Others, however, are of the opinion that one of the 
 structures, being a more stable configuration than the 
 other, really represents the body in the free state, 
 but that this stable form is under certain conditions, 
 and by the action of certain reagents, primarily trans- 
 formed into the less stable, pseudo, or labile form 
 which then undergoes further change. Thus phloro- 
 glucone behaves sometimes as a phenol giving metallic 
 derivatives and methyl ethers, at other times as a 
 
 note, p. 158), by applying spatial considerations to the platinum atom. 
 The latter is regarded as occupying the centre of a regular octahedron, 
 at the six solid angles of which are placed the atoms and radicles 
 constituting the molecule. 
 
1 84 
 
 LAW AND THEORY IN CHEMISTRY 
 
 ketone or carboxyl compound giving oximes. The 
 formula appertaining to its stable phenolic form, i.e., 
 to phloroglucone properly so-called, is 
 
 C.OH 
 HC / \ CH 
 
 HO.C 
 
 C.OH 
 
 CH 
 
 while its ephemeral labile ketonic form demands the 
 structure 
 
 CO 
 
 H 2 C 
 
 OC 
 
 CH 2 
 CO 
 
 CH, 
 
 In some instances, both desmotropic forms are 
 capable of continued and. well differentiated existence 
 in the free state and even in solution. Thus succino- 
 succinic ether exists in both a colourless and a yellow 
 modification. 
 
 It should be remarked that, so far as is known, all 
 cases of tautomerism depend on the mobility in the 
 molecule of hydrogen atoms ; the passage from one 
 desmotropic form to another being effected by the 
 wandering of one or more hydrogen atoms. 
 
 These new views may perhaps be adduced to account 
 
MOLECULAR ARCHITECTURE 185 
 
 for the curious transformations exhibited by cyanogen 
 compounds. Potassium cyanate NOOK prepared from 
 cyanic acid NCOS, when treated with ethyl iodide 
 C 2 H 5 I gives ethyl isocyanate OCNC 2 H 5 , and not ethyl 
 cyanate as would naturally be expected. Yet isocyanic 
 acid OCNH, the desmotropic form of cyanic acid 
 NCOH, is not known in the free state ; the ethyl 
 radicle is required to confer stability on the apparently 
 extremely labile configuration OCNH. 
 
 Similar considerations may be applied in explanation 
 of the behaviour of cyanuric and thiocyanic acids. 
 
CHAPTER VII 
 CHEMICAL EQUILIBRIUMS 
 
 ALTHOUGH the occurrence of chemical interaction 
 between solutions of sodium sulphate and hydrochloric 
 acid is not manifest to the unaided senses by reason of 
 the fact that all the factors and products of the inter- 
 action are soluble colourless bodies, yet we have con- 
 vincing indirect proofs that such an interaction does 
 actually take place. 2 Moreover, the evidence in favour 
 of the occurrence of chemical change between solutions 
 of sodium sulphate and hydrochloric acid is of such a 
 nature that we are forced to conclude that the interac- 
 tion is not correctly represented by the equation 
 
 q + 2HClAq = 2NaCl Aq + H 2 S0 4 Aq. 
 
 For, by convention, this equation when interpreted 
 into words would run : Dilute solutions of sodium 
 
 1 As dissociation phenomena are generally fully treated in elementary 
 text-books, this chapter has been almost wholly devoted to a study 
 of those equilibria which are established in solutions of chemically 
 interacting substances at ordinary temperatures. 
 
 2 See note, p. 193. It is very easy to prove the occurrence of 
 chemical change in the particular system MgSC^Aq + 2NaClAq. 
 although the interaction is not accompanied by visible changes under 
 ordinary conditions. It is only necessary to cool a mixture of the solu- 
 tions of the two salts, when the very slightly soluble sodium sulphate 
 
 will partially crystallise out in a hydrated form. 
 
 186 
 
CHEMICAL EQUILIBKIUM 187 
 
 sulphate and hydrochloric acid when brought together 
 in equivalent quantities 1 completely decompose each 
 other with the formation of equivalent quantities of 
 sodium chloride and sulphuric acid which both remain 
 in solution. 
 
 Now, as a matter of fact, such a complete decomposi- 
 tion of the sodium sulphate and hydrochloric acid does 
 not take place under the specified conditions (dilute 
 solution, and equivalent quantities of the reagents). 
 Only a portion of the system Na 2 S0 4 + 2HC1 + Aq is 
 changed into the system 2NaCl + H 2 S0 4 + Aq. This 
 fact might be adequately represented in the following 
 manner 
 
 (m + n) Na 2 S0 4 Aq + (ra + n) H 2 Cl 2 Aq = m Na 2 Cl 2 Aq + 
 m H 2 S0 4 Aq + n Na 2 S0 4 Aq + n H 2 Cl 2 Aq 
 
 From this representation we learn that when (m -f n) 
 
 1 Those quantities of acids which neutralise a fixed quantity of 
 any base are equivalent. Conversely those quantities of bases which 
 neutralise a fixed quantity of any acid are equivalent. This was the 
 original and most obvious meaning of equivalency, but the original 
 meaning of the term has been widened so that it now applies to salts 
 as well as to acids and bases. A quantity, x grs., of a salt (MA) is 
 said to be equivalent to a quantity y grs. of an acid A', if the quantity 
 of acid A necessary to produce x grs. of the salt MA is equivalent in 
 the original acceptation of the word to y grs. of the acid A'. Thus the 
 molecular weight of sodium sulphate interpreted in grams [142 grs.] is 
 formed from the molecular weight in grams [98 grs.] of sulphuric acid. 
 But 98 grs. of sulphuric acid is equivalent to, i.e., will neutralise the 
 same quantity of any base as, 73 grs. of hydrochloric acid or two molecu- 
 lar weights of hydrochloric acid expressed in grams. Therefore in 
 accordance with the widened significance of the term equivalence now 
 in vogue, 2HC1, HoSC>4, Na 2 S04 represent equivalent quantities of the 
 substances formulated no matter what mass units be used. 
 
i88 LAW AND THEORY IN CHEMISTRY - . 
 
 equivalents of sodium sulphate are brought together 
 with an equal number of equivalents of hydrochloric acid 
 in dilute aqueous solution, only m equivalents of the 
 two substances undergo chemical change, the residual 
 n equivalents of each substance remaining unchanged. 
 But this is not a perfectly satisfactory method of repre- 
 sentation, for this reason. It suggests that the mole- 
 cules of each of the reacting substances must differ in 
 some way or other among themselves, seeing that some 
 enter into chemical reaction under the conditions, while 
 others apparently do not. And this is not in con- 
 formity with the fundamental principle of the atomic- 
 molecular theory, which principle asserts that the 
 molecules of a substance are all precise replicas one 
 of another, both in structure and attributes. If one 
 molecule of sodium sulphate reacts under certain con- 
 ditions, then we should expect all the molecules of 
 sodium sulphate to react under these conditions. 
 
 The difficulty disappears at once if we regard re- 
 actions like the one under consideration as the result of 
 two independent and antagonistic chemical changes 
 proceeding simultaneously within the system. 1 
 
 When hydrochloric acid is added to an equivalent 
 quantity of sodium sulphate in dilute aqueous solution, 
 
 1 That two or more reactions can proceed within a system simul- 
 taneously and independently one of another is the enunciation of the 
 principle of the coexistence of reactions a principle which bears much 
 the same relation to chemistry as does Newton's second law of motion 
 to dynamics. This principle has never been directly proved ; but 
 reasoning founded on the assumption of such mutual independence of 
 reactions has led to results in harmony with experimental fact, and 
 thus the truth of the principle is established. 
 
CHEMICAL EQUILIBK1UM 189 
 
 molecules of sodium chloride and sulphuric acid are 
 formed at a gradually diminishing velocity as time goes 
 on. 1 The new molecules thus formed do not, however, 
 remain inert, but interact to reproduce molecules of the 
 original substances at a gradually increasing velocity. 
 Let us call the passage from Na 2 S0 4 Aq + H 2 Cl 2 Aq to 
 Na 2 Cl 2 Aq + H 2 S0 4 Aq the direct partial change, and 
 the passage in the opposite direction the reverse partial 
 change. 
 
 What causes the velocities of the two changes thus 
 defined to vary from moment to moment ? Guldberg 
 and Waage first definitely answered this question, by as- 
 serting that the velocity of every change at any moment 
 varies directly as the product of the number of the 
 equivalents of the factors of the change present in unit 
 volume of the medium of change. As this number is 
 continually varying, owing to the occurrence of change, 
 it follows that the velocity must also continually vary. 
 
 1 No chemical change is instantaneous, but every change requires for 
 its occurrence a period of time varying from a fraction of a second too 
 small to measure, to several years. The reactions between various acids 
 and alkalis (i.e., neutralisation) take place in so short a time that it has 
 not been possible to measure their absolute durations. Estimates of 
 their relative durations have, however, been indirectly obtained. On 
 the other hand, many of the reactions between organic bodies proceed 
 very slowly indeed. Witness in this connection the slowness of the 
 chemical changes occurring in the so-called ageing of wine. The use 
 of the word velocity in connection with chemical reactions as introduced 
 by Wenzel in 1777 demands explanation. The original dynamical 
 meaning of the term is space passed over per unit time. But the idea 
 of space does not enter into the definition of the term in its chemical 
 acceptation. The velocity of a chemical reaction at any moment 
 simply denotes the quantity of substance (expressed in equivalents) that 
 undergoes change per unit of time at that moment. 
 
IQO LAW AND THEORY IN CHEMISTRY 
 
 Suppose the two substances A and B combine com- 
 pletely to form a third body C. Let us bring together 
 solutions of A and B, such that 2p equivalents of A 
 and 2q equivalents of B are contained in unit volume 
 of the resulting solution. Then the velocity, V, of the 
 reaction (i.e., the rate at which C is formed) will vary as 
 2p x 2q. Or stated algebraically 
 
 Voo pq 
 V = 
 
 Where K is a constant called the velocity constant for 
 the change. 1 After a certain time there will be left only 
 p equivalents of A, and q equivalents of B in unit 
 volume of the medium of change. The velocity of 
 formation of C at this epoch, according to Guldberg 
 and Waage, would be ~Kpq only Jth its original value. 
 The principle here illustrated is often referred to as 
 Guldberg and Waage's law of mass action. 2 
 
 To return now to the interaction of hydrochloric 
 acid and sodium sulphate. The number of equivalents 
 
 1 The velocity constant K must be carefully distinguished from the 
 velocity V. Under given conditions K is constant for any given change, 
 but V of course varies from moment to moment. 
 
 2 The first to prove the influence of mass in chemical changes was 
 Wenzel, 1777. Wenzel, however, confined his investigations to a single 
 class of changes, viz., the interaction of metals with acids of various 
 strengths. In the beginning of this century, Berthollet asserted the 
 influence of mass in chemical changes in general terms. "Every 
 chemical change is essentially of the nature of a combination, and the 
 power of any substance to enter into combination is proportional to its 
 affinity and to its mass." This is practically the text of Berthollet's 
 famous work, Essai de Statique Chimique. According to Berthollet, a 
 small affinity can be compensated by a large mass, and hence complete 
 reactions are the exception and not the rule in chemistry. Suppose 
 
CHEMICAL EQUILIBRIUM 191 
 
 expressed in grams or otherwise (i.e., the masses) of 
 NaCl and H 2 S0 4 resulting from the direct change 
 depends not only on the velocity constant peculiar to 
 the change, but also on the numbers of the equiva- 
 lents (i.e., the masses) of HC1 and Na 2 S0 4 in unit 
 volume of the solution. As these latter numbers are 
 continually decreasing, the velocity of the direct partial 
 change must continually decrease from a maximum value 
 downwards. But as the numbers of the equivalents 
 per unit volume of the factors of the direct partial 
 change decrease, the numbers of the equivalents of 
 the factors of the inverse partial change pari passu 
 increase, consequently the velocity of this change is 
 continually increasing from zero upwards. Hence it 
 necessarily follows that a state of what has been 
 called mobile equilibrium between the direct and re- 
 verse partial changes must finally be reached, whereat 
 the velocities of the two changes being exactly equal, 
 
 the body C has a stronger affinity for A than has B. Then if C be 
 added to AB in equivalent quantity, the reaction 
 
 AB + C = AC + B 
 
 will proceed, but not to completion, for the gradually increasing mass 
 of B will eventually compensate its smaller affinity, and make it for- 
 midable enough to contest successfully the diminishing quantities of C. 
 Before Berthollet's time it had been held that reactions are determined 
 entirely by the stronger affinities ; to which view the fdea of incomplete 
 reactions was quite foreign. From his fundamental premiss, however, 
 Berthollet made deductions which Proust proved to be incorrect (see 
 p. 74), and this overthrow of the incorrect deductions unfortunately 
 brought the correct premiss from which they were derived into dis- 
 repute and temporary oblivion. Guldberg and Waage's law of mass 
 action is nothing else than Berthollet's views reinstated and thrown 
 into mathematically exact form. 
 
192 LAW AND THEORY IN CHEMISTRY 
 
 effective chemical change one way or the other ceases. 
 When this permanent equilibrium point is arrived 
 at, as many equivalents of HC1 and Na 2 S0 4 are de- 
 composed by the direct partial change per unit time 
 as are formed by the reverse partial change in the 
 same time unit ; but the molecules of Na 2 S0 4 , HC1, 
 -IIN0 3 , and NaCl existing at one epoch are not 
 necessarily the same ones which will be present at 
 another subsequent epoch. If they do chance to be 
 identically the same molecules at the two epochs, we 
 can assert that they have not had a continued exist- 
 ence in the interim. 
 
 In the representation of these so-called incomplete 
 reactions, wherein a state of equilibrium is finally 
 reached, van't Hoff has proposed to replace the sign 
 of equality in the ordinary equations by a pair of 
 oppositely directed arrows, thus 
 
 H 2 Cl 2 Aq + Na 2 S0 4 Aq ^ Na 2 Cl 2 Aq + H 2 S0 4 Aq. 
 
 Now, thoroughly appreciating the nature of the 
 dynamical equilibrium which marks the cessation, not 
 of chemical change, but of effective chemical change, 
 in these incomplete reactions, we will return to that 
 method of their representation which we first gave, 
 and which for the immediate purpose in hand is more 
 convenient than the one proposed by van't Hoff. 
 
 Let us suppose that in the general equation 
 (m + ra)Na 2 S0 4 Aq + (m 
 raH 2 S0 4 Aq 
 
 m + n equals 100. The question arises, what are the 
 
CHEMICAL EQUILIBRIUM 193 
 
 individual values of m and n ? This question has been 
 submitted to experiment, and it has been ascertained 
 that m and n in the above reaction have the approxi- 
 mate values 66f and 33J respectively. 1 Moreover, 
 exactly the same state of final equilibrium is arrived 
 at if, instead of setting out with hydrochloric acid and 
 sodium sulphate, we start with hydrochloric acid and 
 the proximate constituents of sodium sulphate, viz., 
 caustic soda and sulphuric acid, in accordance with 
 the following equation 
 
 (m + n) Na 2 2 H 2 Aq + (m + n) H 2 S0 4 Aq + (m + n) 
 H 2 Cl 2 Aq = mNa 2 C! 2 Aq + mH 2 S0 4 Aq + 
 
 This latter method of regarding the reacting system 
 emphasises the fact that the reaction virtually consists 
 in a strife between two equivalent quantities of acids 
 to appropriate a quantity of base only just sufficient 
 to neutralise either acid separately. As a result of 
 this competition the base divides itself between the 
 two acids, twice as much base (i.e., 66f equivalents) 
 combining with the hydrochloric acid as combines with 
 the sulphuric acid (33J- equivalents). 
 
 In accordance with the results of such investigations 
 as the one just described, acids are classed as weak or 
 strong. When two acids in equivalent quantities and 
 
 1 For the actual carrying out of methods suited to the investigation 
 of equilibria see Muir and Carnegie's Practical Chemistry, pp. 141-146, 
 176-184. 
 
194 LAW AND THEORY IN CHEMISTRY 
 
 in dilute aqueous solution are presented to a dilute 
 aqueous solution of a quantity of base just sufficient 
 to neutralise either acid separately (or what amounts to 
 the same thing, when an acid is allowed to react with 
 an equivalent quantity of the salt of another acid in 
 dilute aqueous solution) the base is in general divided 
 between the two acids, and the strengths (or affinities, 
 or avidities) of the two acids are proportionate to the 
 quantities of base each appropriates. Thus it follows 
 that sulphuric acid is a weaker acid than hydrochloric, 
 the latter acid being approximately twice as strong as 
 the former. By conducting numerous quantitative ex- 
 periments, involving different pairs of acids, we can 
 obtain a table of the relative affinities or strengths of 
 the acids, such as the following one, in which, however, 
 the values assigned the various acids are admittedly 
 only approximate : 
 
 Nitric acid . . . . . TOO 
 
 Hydrochloric acid . . . . 1 -00 
 
 Hydrobromic acid .... '89 
 
 Hydriodic acid ..... -70 
 
 Sulphuric acid ..... '49 
 
 Selenic acid . . . . . '45 
 
 Trichloracetic acid . . . '36 
 
 Orthophosphoric acid . . '25 
 
 Oxalic acid . . . . '24 
 
 Hydrofluoric acid . . . . '05 
 
 Citric acid . . . . . '05 
 
 Acetic acid ..... '03 
 
 Boric acid -01 
 
 Silicic acid '00 
 
CHEMICAL EQUILIBRIUM 195 
 
 As will be seen nitric and hydrochloric acids are the 
 two strongest acids, being approximately equal in 
 strength. Compared with either of these, acetic acid 
 is seen to be a very feeble one. If equivalent quan- 
 tities of hydrochloric acid and acetic acid were added 
 to a dilute aqueous solution of caustic soda, contain- 
 ing just enough caustic soda to neutralise either acid 
 separately, the caustic soda would divide itself between 
 the hydrochloric and acetic acids in the ratio of 1 : 03, 
 i.e., about 2'91 per cent, of the soda would combine with 
 the acetic acid and 97'09 per cent, with the hydrochloric 
 acid. In a precisely similar way bases have been classed 
 as weak or strong. Of a pair of bases presented in 
 dilute aqueous solution to an acid only just sufficient 
 to neutralise either base separately, that base which 
 appropriates, or combines with, the greater part of the 
 acid is the stronger base. The following table gives 
 the relative strengths of a few of the better investi- 
 gated bases in terms of the arbitrary value unity 
 assigned to the strongest base, lithium hydroxide : 
 
 Lithium hydroxide .... 1*00 
 
 Sodium hydroxide .... *98 
 
 Potassium hydroxide . . . . '97 
 
 Thallium hydroxide .... '90 
 
 Piperidin . . . . . '16 
 
 Ammonia ...... *02 
 
 We have seen that hydrochloric acid is approximately 
 twice as strong as sulphuric acid if the criterion be a 
 tug-of-war for a quantity of soda insufficient to meet 
 
196 LAW AND THEORY IN CHEMISTRY 
 
 the conjoint demands of the two acids. It has been 
 found that approximately the same relation holds 
 between the strengths of the two acids whatever base 
 be made the object of competition, whether potash, or 
 soda, or lime, &c. Moreover, as is well known, acids 
 have the power of inverting cane sugar, that is, of 
 transforming it in the presence of water into a mixture 
 of dextrose and laevulose ; and also in this transforma- 
 tion hydrochloric acid is found to act twice as ener- 
 getically as sulphuric acid. Numerous other reactions 
 might be adduced which are either brought about or 
 accelerated by acids, and in all of them hydrochloric 
 acid is approximately twice as effective as sulphuric 
 acid. 
 
 Similar remarks apply to changes induced or acceler- 
 ated by bases. If lithium hydroxide is fifty times as 
 strong as ammonia when the struggle for a particular 
 acid is the criterion, then it is found to be approxi- 
 mately fifty times as effective as ammonia in bringing 
 about any other change demanding the intervention of 
 a basic substance. 
 
 On the basis of the approximately concordant results 
 for the strengths of acids and bases derived from the 
 investigation of such diverse changes as we have hinted, 
 it was at one time claimed that the values thus assigned 
 to the acids and bases as representing their strengths, 
 could they be freed from all experimental error, would 
 be perfectly characteristic numbers for the respective 
 acids and bases, quantitatively conditioning all the 
 reactions brought about by them. In pursuance of 
 
CHEMICAL EQUILIBRIUM 197 
 
 these views, it was proposed to give the generic name 
 of specific affinity constants to the numbers represent- 
 ing the strengths of the acids and bases. 
 
 On fuller investigation, however, it appeared that 
 neither the claim nor the proposal could be admitted; 
 for the partition of a base between two acids depends 
 not only on the nature of the contending acids, but 
 also on the quantity of water present. A table of 
 strengths or affinities constructed from the results of 
 experiments on the partition of a base between acids 
 in normal solution, 1 does not present the acids in 
 the same order as a table founded on the results of 
 investigations of deci-normal acid solutions. In other 
 words, water cannot be regarded as a merely passive 
 medium when the power of acids is in question; "the 
 power of an acid to do " is a function of the state of 
 dilution of the acid. For instance, in the system acid 
 A, acid A', "base, water, we have not merely a tug-of- 
 war of the two acids A and A' for the base, but the 
 final equilibrium is the expression of the resultant of 
 (1) the "affinity" of A for water, (2) the "affinity" 
 of A' for water, (3) the "affinity" of the base for 
 water, (4) the "affinity" of A for the base, and 
 (5) the "affinity" of A' for the base; and in con- 
 
 1 Let the numerical value of the quotient 
 molecular weight of acid 
 
 number of replaceable H atoms in molecule 
 
 be called n. Then a normal solution of an acid contains n grams of 
 the pure acid per litre of solution at a specified temperature. A deci- 
 normal solution has T \th the concentration of a normal one. 
 
198 LAW AND THEORY IN CHEMISTRY 
 
 formity with the law of mass action, every alteration 
 in the quantity of water present necessarily alters 
 the final distribution of matter marking the state of 
 equilibrium from which the strengths of the acids are 
 judged. 
 
 Quite recently, however, it has been found possible 
 to obtain for the acids characteristic numbers which 
 are independent of their greater or less dilution, and 
 of the particular reactions in which the acids may 
 be implicated. These numbers are now regarded as 
 the true affinity constants of the acids, but they 
 necessarily have quite a different significance from the 
 old system of supposed constants derived from the 
 study of the partition of acids between bases, &c., 
 seeing that they are independent of dilution. 
 
 When an acid is dissolved in a large quantity of 
 water, it behaves in many respects as if it were (and 
 is accordingly by some believed to be) partially decom- 
 posed into its ions (see note, p. 153). Thus a dilute 
 solution of hydrochloric acid behaves as if it contained, 
 in addition to molecules of HC1, ions of hydrogen and 
 chlorine. If at a certain dilution K the number of 
 undecomposed molecules is just equal to the number 
 decomposed into ions, then K is taken as a measure of 
 the true specific affinity constant of the acid. 
 
 The investigation of affinity in accordance with these 
 conceptions has as yet been chiefly confined to organic 
 acids, and it is found that the order of the acids in a 
 table of affinities, according to the new definition of 
 affinity, is practically the same as the average order in 
 
CHEMICAL EQUILIBRIUM 199 
 
 the old tables of supposed affinity constants which 
 were derived from the results of different reactions 
 brought about by acids of varying concentration. 1 
 
 We have seen (page 57) that certain phenomena 
 exhibited by fairly strong aqueous solutions point to 
 definite combinations between water molecules and 
 molecules of dissolved substance, i.e., to the existence 
 in solutions of definite liquid hydrates. It is supposed 
 that these liquid hydrates are very unstable bodies 
 which at ordinary temperatures are partially dis- 
 sociated, 2 the original hydrates forming with the 
 products of their dissociations mobile equilibria as 
 represented in the following equations : 
 
 m (An H 2 0) ^ mn H 2 + ra A 
 
 m' (An' H 2 0) ; m'ri H 2 + m r A &c., 
 
 where (m + m + &c.) A represents the salt dissolved 
 in mn = m'ri = &c. molecules of water. The inter- 
 pretation of the above symbolism is that in a solution 
 of A of determined concentration, the hydrates AnHf), 
 A%'H 2 &c. are present, but that these hydrates 
 
 1 In the algebraical expression through which the desired affinity 
 constants are obtained from the results of experiment, there is an 
 inherent weakness which makes itself especially felt in the case of 
 the stronger inorganic acids. For further details on the subject 
 of affinity constants the reader is referred to the Lehrbuch der Attge- 
 meincn Chemie of Ostwald, to whom we owe much of the best work 
 that has been done in the direction of raising affinity to a quantitative 
 dignity. 
 
 2 A dissociation is a reversible decomposition. When KClOs is 
 heated it decomposes into KC1 and 30, for these two substances do not 
 recombine on cooling to reform potassium chlorate. When PCls is 
 heated, it dissociates into PC1 3 and Cla, because these two gases as 
 they cool in contact combine together completely to reform PC1 5 . 
 
200 LAW AND THEOKY IN CHEMISTKY 
 
 partially dissociate into anhydrous salt, 1 (mA, m'A &c.) 
 and water (m?iH 2 0, mVH 2 &c.) till mobile equilibria 
 are established for each system, i.e., until in each system 
 as many molecules of the hydrate are dissociated as are 
 reformed per unit of time. As the concentration of 
 the solution is varied, the ratios in which the various 
 dissociating hydrates are present also varies, and at 
 certain stages in the concentration new dissociating 
 hydrate systems would make their appearance, while 
 pre-existing systems would disappear. According to 
 the views here illustrated, solutions have been defined 
 as " fluid and unstable, but definite chemical compounds 
 in a state of dissociation." 
 
 All the incomplete reactions we have already con- 
 sidered have been homogeneous, 2 but homogeneity in 
 
 1 Or it may be, into a hydrate of lower hydration in accordance 
 with the equation 
 
 m(AnH 2 0)l;m(A^H 2 o) + (mn- ?\H Z 0, 
 
 where TO, n, and x are integers. 
 
 2 When all the products arid factors of a reaction are in the same 
 physical state (same state of aggregation), the reaction is said to be 
 homogeneous. The following are examples of homogeneous reactions : 
 
 C1 2 + H 2 = 2HC1. NaCl Aq + KN0 3 Aq - KC1 Aq + NaNO 3 Aq. 
 
 gases. solutions. 
 
 Fe + S = FeS. 
 
 solids. 
 
 The combination of gaseous S0 2 with solid Pb0 2 to form solid 
 PbSO4 is illustrative of non-homogeneous reactions. The significance 
 of the terms homogeneous and non-homogeneous used in connection 
 with reactions or systems must be carefully distinguished from the 
 significance of the same terms when used, as in Chap. III., in con- 
 nection with substances. 
 
CHEMICAL EQUILIBRIUM 201 
 
 this sense is not an essential condition of incomplete 
 reaction. The very familiar reaction between solutions 
 of common salt and silver nitrate is not correctly repre- 
 sented in the equation 
 
 AgNOgAq + NaCl Aq = AgCl + NaN0 8 Aq, 
 
 for when effective chemical change between strictly 
 equivalent quantities of AgN0 3 and NaCl has ceased, 
 it is easy to prove the presence in the non-homogeneous 
 system of traces of undecomposed AgN0 3 and NaCl. 
 In other words, a condition of equilibrium finally 
 supervenes, the equilibrium in this case, however, being 
 so much in favour of the direct partial change, that 
 unless for very accurate work, the decomposition of 
 AgN0 3 by an equivalent quantity of NaCl may be 
 regarded as practically complete. 
 
 The interaction of steam and heated iron is another 
 familiar instance of an incomplete reaction taking place 
 in a non-homogeneous system. 
 
 3Fe + 4H 2 ^ Fe 3 4 + 4H 2 
 (solid) (gas) (solid) (gas) 
 
 The sole condition for incompleteness of reaction 
 between equivalent quantities of interacting substances 
 resulting in mobile equilibria is that the products of 
 the change must be of such a nature that they are all 
 retained within the sphere of action of the system 
 considered. 1 If by any means removal from the sphere 
 
 1 It follows that the salts in solution in mineral waters, &c., must 
 constitute very complicated dynamic equilibria. The schemes in which 
 analysts are wont to express the results of their analyses are mislead- 
 
202 LAW AND THEORY IN CHEMISTRY 
 
 of action of one or more of the products of the change 
 is effected, the reaction proceeds to completion. Thus, 
 if in the interaction of steam and heated iron it were 
 possible to allow the hydrogen to escape as soon as 
 formed, without at the same time permitting any steam 
 to escape, then three atoms of iron would completely 
 decompose four molecules of steam. One of the factors 
 necessary to the reverse partial change being removed, 
 that change cannot take place, and the direct partial 
 change meeting with no opposition proceeds to comple- 
 tion. To take another instance actually realised by 
 Berthelot and St. Gilles. When benzoic acid and 
 alcohol are brought together they slowly interact with 
 the production of benzoic ether (ethyl benzoate) and 
 water, which in turn simultaneously interact, reforming 
 the original factors of the change. 
 
 C r) H 5 COOH + C 2 H 5 OH ^ C 6 H 5 COOC 2 H 5 + H 2 0. 
 
 Under ordinary conditions effective chemical change 
 ceases when about 66 per cent, of the benzoic acid and 
 alcohol have been transformed into benzoic ether. But 
 when the water resulting from the change is removed 
 from the sphere of action, the whole of the benzoic acid 
 and alcohol is forthwith etherified. This removal of 
 the water from the sphere of action is effected by the 
 addition of barium oxide to the system. As soon as 
 
 ing. The bases in a homogeneous system can be accurately determined 
 and also the acids, but it is impossible from the results of the quanti- 
 tative analysis only to apportion the acids to the bases so as to represent 
 the actually existing constitution of the system. 
 
CHEMICAL EQUILIBRIUM 203 
 
 any water is formed it combines with the oxide forming 
 inert barium hydroxide. 
 
 It is to be noted that the same state of equili- 
 brium is reached whether we set out with the system 
 C 6 H 5 COOH + C 2 H 5 OH or with the systemC 6 H 5 COOC 2 H 5 
 -f- H 2 0. In the former case effective change proceeds 
 till 66 per cent, of the benzoic acid and alcohol have 
 been transformed into benzoic ether and water ; in the 
 latter case effective change proceeds until 66 per cent, 
 of the ether and water remain undecomposed. This 
 is a perfectly general characteristic of these mobile 
 equilibria. Eepresenting incomplete reactions by the 
 general equation 
 
 AB + A'B' : A'B + AB' 
 
 (1) (2) 
 
 we may say that in all cases the distribution of 
 matter which obtains when equilibrium is established 
 is independent of whether the point of departure be 
 made the system (1) or (2). In a few particular cases 
 this statement admits of very simple proof. For 
 instance, a mixture of equivalent quantities of KClAq 
 and NaN0 3 Aq gives exactly the diffusate as a mixture 
 of equivalent quantities of KN0 3 Aq and NaClAq, 
 showing that in both cases the same equilibriated 
 system 
 
 zNaClAq + asKNOgAq + (1 - z)KC!Aq + 
 .(1 - 
 
 is undergoing dialysis. 
 
204 LAW AND THEORY IN CHEMISTRY 
 
 Another method of rendering an incomplete reaction 
 practically complete as regards some one or more of 
 the members constituting the system will at once be 
 evident on referring to Guldberg and Waage's law of 
 mass action. This method consists in largely increas- 
 ing the mass of one of the factors of the change rela- 
 tively to the other. Thus, if in the etherification of 
 alcohol by benzoic acid we largely increase the number 
 of equivalents of alcohol relatively to the number of 
 equivalents of benzoic acid, or vice versd, then, in either 
 case, we approach indefinitely close to a state of com- 
 plete etherification of the acid on the one hand, or the 
 alcohol on the other. 
 
 It is in virtue of the action of mass that we are 
 able to completely convert NaCl into NaN0 3 by means 
 of HN0 3 Aq, or conversely to change NaN0 3 into NaCl 
 by means of HClAq. When NaCl, for instance, is 
 treated with HN0 3 Aq a partial change takes place, and 
 NaCl, HC1, NaN0 3 , and HN0 3 are all present. If the 
 system be now heated, the free acids owing to their 
 volatilities pass away, leaving non-volatile NaCl mixed 
 with a little non-volatile NaN0 3 . If more HN0 3 Aq is 
 added to this residue, and the system again heated, a 
 second residue richer in NaN0 3 will result ; and so 
 by repeating the operations of adding HN0 3 Aq and 
 evaporating often enough (i.e., by using a large enough 
 relative mass of HN0 3 ), the NaCl can be completely 
 changed into NaN0 3 . The case of the converse trans- 
 formation of NaN0 3 into NaCl by means of a large 
 excess of HC1 is quite similar. 
 
CHEMICAL EQUILIBRIUM 205 
 
 In the light of the facts just treated, a reference to 
 the remarks made on solution both in this chapter and 
 in Chapter III. leads us into difficulties. 
 
 If a substance in solution be largely diluted, i.e., if 
 the relative mass of water be largely increased, we 
 should expect, in accordance with what precedes, that 
 the excess of water would lead to a great increase in 
 the amounts and stabilities of the higher hydrates of 
 the dissolved substance. But those who have experi- 
 mented most with very dilute solutions, find that their 
 results are best summarised in terms of the hypothesis 
 that hydrates do not exist at all in dilute solutions. 
 It appears as if the increased dilution not only breaks 
 up the molecular complexes of water molecules and 
 salt molecules constituting the hydrates, but actually 
 destroys the integrity of the salt molecules themselves, 
 resolving these into their ions. 1 For the present, until 
 more light is shed on the vexed question of solution, 
 we must be content to remember that this difficulty 
 exists without attempting to decide it one way or the 
 other. 
 
 The question now arises for discussion, are all re- 
 actions between equivalent quantities of mutually 
 interacting bodies incomplete ? May not one or more 
 of the products of certain changes be of such a nature 
 
 1 See note, p. 153. An ion may consist of a group of atoms, e.g., 
 SO.*, or of a single atom, e.g., Na. In the latter case the ion and the 
 atom are supposed not to be identical. For the present it is customary 
 to attribute the assumed difference between them to the possession by 
 the ion of a certain charge of electricity proportional to the valency of 
 the atom. 
 
206 LAW AND THEORY IN CHEMISTRY 
 
 that they are removed from the system's sphere of 
 action as quickly as they are formed, in virtue of their 
 own peculiar properties ? Changes giving rise to 
 bodies with such properties, would proceed to comple- 
 tion even although their factors were brought together 
 in strictly equivalent quantities. 
 
 Stas has shown that AgBr is not acted on at all by 
 a solution of NaN0 3 , or that if there be any action it 
 is too slight to be detected by the analytic means at 
 our disposal. Hence it follows that the reaction between 
 AgN0 3 Aq and NaBrAq is, so far as we can tell, a 
 complete one, and that the equation 
 
 AgN0 3 Aq + NaBrAq = AgBr + NaN0 3 Aq 
 
 is experimentally realised. 
 
 There are many well-known reactions, such as that 
 between H 2 S0 4 Aq and BaCl 9 Aq, viz., 
 
 H 2 S0 4 Aq + BaCl 2 Aq = BaS0 4 + 2HClAq, 
 
 which are for practical purposes regarded as complete, 
 but which, as a matter of fact, are incomplete reactions 
 in which the distribution of matter marking the attain- 
 ment of equilibrium is overwhelmingly in favour of the 
 direct partial change. The BaS0 4 , formed in the above 
 reaction, is slightly soluble in (i.e., interacts with) 
 dilute solutions of HC1, and, in virtue of this slight 
 solubility, there is a reverse partial change, which is 
 the condition for incompleteness of reaction and the 
 establishment of an equilibrium. But the reverse 
 partial change in this case is of very small moment. 
 
CHEMICAL EQUILIBRIUM 207 
 
 The subject of neutralisation calls for a word or two 
 here. Those particular and important branches of 
 volumetric analysis known as alkalimetry and acidi- 
 metry are founded on the assumption of the absolute 
 truth of such equations as 
 
 H 2 S0 4 Aq + 2NH 4 OHAq = (NH 4 ) 2 S0 4 Aq + 2H 2 
 HClAq f NaOHAq = NaClAq + H 2 0. 
 
 That is to say, these reactions are regarded as complete 
 although there is evidently no removal from the sphere 
 of action, artificial or otherwise, of the products of the 
 neutralisations of equivalent quantities of acids and 
 bases. Here, again, it is probable that we have in- 
 stances of incomplete reactions in which the reverse 
 partial change is so insignificant as to be negligible. 
 That ammonium sulphate is slightly decomposed by 
 water with the formation of free acid and ammonia is 
 probable from the fact that air when passed through 
 solutions of (NH 4 ) 2 S0 4 acquires alkaline properties 
 the solutions themselves meanwhile developing acid 
 characters. This behaviour of ammonium sulphate 
 solution has been adduced to explain the completeness 
 of the interaction of equivalent quantities of ammo- 
 nium sulphate (NH 4 ) 2 S0 4 Aq and potassium carbonate 
 K 2 C0 3 Aq. Seeing that both the factors of this inter- 
 action, and all the possible products formed by double 
 decomposition, are soluble in water and therefore remain 
 within the sphere of action, one would be inclined, 
 off-hand, to class this particular change with the in- 
 complete reactions. 
 
2o8 LAW AND THEORY IN CHEMISTRY 
 
 According to Berthelot the completeness of the inter- 
 action finds an explanation in terms of the partial 
 decomposition of one of the interacting factors. The 
 ammonium sulphate and water may be regarded as 
 giving rise to some such equilibrium as is represented 
 in the following equation 
 
 (NH 4 ) 2 S0 4 Aq + 2H 2 OAq ^ 2NH 4 OHAq + H 2 S0 4 Aq. 
 
 The potassium carbonate destroys this equilibrium by 
 neutralising (and so removing from the sphere of action) 
 the free sulphuric acid present at any moment ; more 
 ammonium salt reacts with water in the tendency to 
 again establish the primitive equilibrium, and so the 
 cycle repeats itself until practically the whole of the 
 ammonium sulphate is decomposed. 
 
 The experimental investigation of the possibility of 
 the establishment of equilibria in the cases of reactions 
 evolving gases (which under ordinary circumstances 
 escape from the sphere of action) is a very difficult 
 matter. A definite answer has as yet not been furnished 
 to the question, is the equation 
 
 Zn + E 2 S0 4 Aq 1; ZnS0 4 Aq + H 
 
 2 , 
 
 a correct representation of facts, provided the hydrogen 
 evolved is prevented from leaving the system would 
 the reaction under such conditions be found to be 
 actually incomplete? In the meantime one has little 
 hesitation in giving a theoretical answer in the affirma- 
 tive to such a question. 
 
 The existence of such equilibria as we have been 
 
CHEMICAL EQUILIBEIUM 209 
 
 discussing disproves a principle which was enunciated 
 by Thomsen in 1854*, and again in 1864 by Berthelot. 
 The latter naturalist has done so much work in the 
 attempt to establish the principle on an experimental 
 basis that it is now generally referred to as Berthelot's 
 law of maximum work, although priority in the matter 
 undoubtedly belongs to Thomsen. 1 Berthelot enunci- 
 ated his " law " in the following terms : 
 
 Every chemical change realised without the intervention 
 of external energy tends to the formation of that body or 
 system of bodies, the production of which is accompanied 
 by the development of the greatest quantity of heat. 
 
 This statement, which simply asserts that every 
 chemical reaction tends to make the system assume 
 that state, in the attainment of which it liberates most 
 heat, further implied to Berthelot the necessity for the 
 occurrence of every transformation that would involve 
 an evolution of heat, and the impossibility of the 
 spontaneous occurrence of every transformation that 
 would involve an absorption of heat. 2 And it is this 
 
 1 Berthelot called his principle the "Law of maximum work," from 
 its analogy to a well-established principle in mechanical energetics. 
 If a raised stone be allowed to fall, it falls vertically to the earth's 
 surface, and never in a direction inclined to the plumb line. The 
 vertical direction is the one in which occurs a maximum change of 
 potential into kinetic energy per unit time. This is an instance 
 illustrating the so-called maximum principle which asserts that of all 
 possible changes that one will take place which involves the greatest 
 transformation of energy per unit time. (Throughout the remainder 
 of this chapter it is assumed that the reader has an elementary know- 
 ledge of general energetics.) 
 
 2 As a corollary to this "law," Berthelot stated his theorem of the 
 necessity for reactions as follows : " Every chemical change which 
 
 
 
210 LAW AND THEORY IN CHEMISTRY 
 
 practical aspect of the " law " that has chiefly appealed 
 to, and engaged the attention of, chemists generally. 
 
 Suppose we have a substance quantitatively repre- 
 sented by AB, and we wish to know whether it will 
 interact with a third body quantitatively represented 
 by to form either the bodies AC + B or the bodies 
 BC + A. We may provisionally regard such an inter- 
 action as taking place in two consecutive stages ; in 
 the first stage AB is decomposed into A and B. and 
 then in the second stage A combines with B or C, as 
 the case may be. If A and B combine together to 
 form AB with the evolution of a quantity of heat 
 energy h, we know that exactly the same quantity of 
 energy h must be added in order to reverse the change 
 and resolve AB into A and B. 1 Suppose that the 
 combination of A and gives rise to an evolution of 
 heat energy represented by H, and that the combina- 
 tion of B and C is attended by an evolution of heat H'. 
 
 can be accomplished without the aid of a preliminary action or the 
 addition of energy from without the system, necessarily occurs if it is 
 accompanied by disengagement of heat." This theorem seems to have 
 been generally applied without regard to the significance of the some- 
 what indefinite qualifications " which can be accomplished . . . from 
 without the system." 
 
 1 The principle, that the quantity of heat energy absorbed in de- 
 composing a given mass of a compound into its elements is exactly 
 equal to the heat energy evolved when the elements combine to form 
 the given mass of the compound, was first given by Lavoisier and 
 Laplace. This principle, which is a necessary consequence of the 
 more general principle of the conservation of energy, is also true for 
 chemical changes other than mere synthesis and analysis. If the 
 passage of a complex system from any state A to any other state B, 
 by any path whatsoever, absorbs a quantity of heat energy H, then 
 the reverse passage from B to A by any path will be attended with 
 the evolution of H heat units. 
 
CHEMICAL EQUILIBEIUM 211 
 
 The following cases must be considered : 
 (1.) H and H' may both be less than h, in which 
 event the "law," or rather its popularised implication, 
 pronounces against the spontaneous occurrence of either 
 change, seeing that both changes under these conditions 
 would be endothermic. 1 (2.) H and H / may both be 
 greater than h, in which case both reactions are exo- 
 thermic, and therefore possible under the conditions. 
 The relative magnitudes of H and H/ according to the 
 " law " decide as to which of the two possible changes 
 will actually occur. If H > H', then AC + B will be 
 formed exclusively ; if H' > H, then BC + A will be 
 the sole products of the reaction. 
 
 When steam is passed over hot iron, chemical change 
 ensues, resulting in the production of magnetic oxide 
 of iron and hydrogen 
 
 3Fe + 4H 2 = Fe 3 4 + 4H 2 , 
 
 but if retaining the same conditions copper be substituted 
 for iron, the steam is not decomposed with the production 
 of copper oxide and hydrogen in accordance with the 
 
 equation 
 
 Cu + H 2 = CuO + H 2 . 
 
 These facts and many others of a similar nature might 
 be adduced as confirmations of Berthelot's "law." 
 
 1 It is to be regretted that the application of so many different 
 significations to the terms exothermic and endothermic has led many 
 chemists to avoid their use altogether. Rigidly defined they are 
 exceedingly convenient terms for the description of thermo-chemical 
 phenomena. In the text an endothermic reaction is taken to mean 
 any reaction which is accompanied by an absorption of heat ; an exo- 
 thermic reaction, any reaction which is accompanied by an evolution 
 of heat. The terms as thus used have no quantitative significance. 
 
212 LAW AND THEORY IN CHEMISTRY 
 
 For since the formation of a gram molecule (232 grs.) 
 of Fe 3 4 from its elements is accompanied by an 
 evolution of 264,700 heat units, while the resolution 
 of four gram molecules (72 grs.) of steam into its 
 elements is attended by the absorption of only 232,000 
 heat units, it follows that the realisation of the first 
 change will be attended by the evolution of heat energy 
 to the extent of 32,700 units. On the other hand, 
 the 37,200 units of heat evolved when a gram molecule 
 (80 grs.) of copper oxide is formed from its elements 
 is less than that required, 58,000 units, to resolve a 
 gram molecule (18 grs.) of gaseous water into its 
 elements, and so the realisation of the second reaction 
 would involve an absorption of heat. 1 
 
 Further, the generalisation that the readiness and 
 intensity with which reactions take place increase with 
 the thermal values of the reactions, is also in accordance 
 with Berthelot's "law." 
 
 But on the other hand there are very many facts 
 which are at variance with the "law." Suppose we 
 wish to know, without actually trying the experiment, 
 whether hydrochloric acid gas acts on silver in accor- 
 dance with the empirical equation 
 
 HC1 + Ag - AgCl + H. 
 Looking up the heat of formation of HC1 (i.e., the 
 
 1 The possibility that reactions could take place in such way that 
 the hydrogen of the water combines with the metals and the oxygen 
 is set free, is presumed to be precluded by the fact that such changes 
 would be strongly endothermic, though through lack of data the 
 exact values of the heat absorption in each case cannot be given. 
 
CHEMICAL EQUILIBEIUM 213 
 
 heat evolved when a gram molecule of hydrochloric acid 
 gas is formed from its elements), we find it to be 
 220,000. The heat of formation of AgCl is 294,000. 
 Now, applying Berthelot's " law " to these data, we 
 are led to the conclusion that silver will be attacked 
 by hydrochloric acid. But this conclusion is false, for 
 silver is quite unaffected by hydrochloric acid. Again, 
 when electric sparks are sent through a mixture of 2 
 volumes of hydrogen, with 2 volumes of chlorine and 
 1 volume of oxygen, hydrochloric acid is exclusively 
 formed, although the formation of water from the 
 hydrogen and oxygen would be attended by a far 
 greater production of heat. 
 
 But the existence of chemical equilibria is, as we 
 have already hinted, the most generalised argument 
 that can be brought forward in opposition to the 
 "law" of Berthelot. Let us suppose that the direct 
 partial change of a particular case of equilibrium is 
 attended by a heat evolution, then it follows from the 
 well-established principle of the conservation of energy 
 that the reverse partial change must involve a heat 
 absorption. For the same reason, if the direct partial 
 change is endothermic instead of exothermic, then the 
 reverse partial change must be exothermic instead of 
 endothermic. That is to say, an equilibrium demands 
 the occurrence of two simultaneous chemical changes, 
 one of which is associated with an evolution of heat, 
 the other with an absorption of heat ; and when equili- 
 brium is attained, these two changes are proceeding at 
 exactly the same rates. If Berthelot's " law " were true, 
 
2i 4 LAW AND THEOKY IN CHEMISTRY 
 
 all reactions would belong to the complete type, and 
 incomplete reactions, i.e., equilibria, would be impossible. 
 
 If we interpret Thomsen and Berthelot's principle 
 as merely asserting that every purely chemical change 
 of a complete type that takes place spontaneously must 
 be accompanied by a loss of chemical energy which 
 will finally appear as heat the lowest form of energy 
 then we are bound to admit the truth of the prin- 
 ciple at the same time as we deprive it of all practical 
 import. For all the so-called chemical changes with 
 which we have to do are not purely chemical changes, 
 in this sense, that other forms of energy besides 
 chemical energy undergo increase and decrease as the 
 distribution of matter changes, and the heat evolved 
 or absorbed by the reaction is a complex quantity, re- 
 presenting not merely the changes of chemical energy, 
 but the resultant of the sum total of energy changes 
 of all kinds. A. purely chemical change is a fiction of 
 much the same order as an absolutely rigid bar, or a 
 perfectly frictionless constraint. 
 
 Further, it can be shown that in incomplete re- 
 actions (i.e., in cases of equilibrium) the chemical 
 energy of the system suffers no decrease during the 
 course of the reaction ; hence, if purely chemical energy 
 were the only form of energy to be considered, it 
 follows that all reactions leading to the establishment 
 of an equilibrium ought to proceed without any heat 
 absorption or evolution. But this is by no means 
 the case, and so we are forced to the conclusion 
 that in the so-called chemical changes other forms of 
 
CHEMICAL EQUILIBRIUM 215 
 
 energy besides chemical energy (e.g., volume energy, 
 heat energy, surface energy, &c.) must undergo trans- 
 formations in order that the actually observed thermal 
 disturbances may be accounted for. 
 
 But recognising the multiplicity of the specific energy 
 changes which give rise to the thermal phenomena 
 attendant on chemical change, the question may still 
 be put, is there absolutely no connection under any 
 conditions between the value and sign of the thermal 
 changes and the necessity for the occurrence of 
 chemical change ? Cannot we predict under any cir- 
 cumstances which of two or more possible changes will 
 occur ? 
 
 It was a great advance that Horstmann made when 
 he showed that the formulae of thermodynamics can be 
 applied to problems in the domain of chemistry. The 
 leading features of the dissociation of such a substance 
 as chalk, in accordance with the symbolism 
 
 CaC0 8 ; CaO + C0 2 , 
 
 are very similar to those of the evaporation of a homo- 
 geneous liquid such as water. But the application of 
 the second law of thermodynamics to the process of 
 evaporation leads to the establishment of very im- 
 portant relationships expressible in a simple formula, 
 and Horstmann, on the strength of the strong analogy 
 between the processes of evaporation and dissociation, 
 applied this particular formula to certain cases of dis- 
 sociation. The results to which the formula led were 
 found to be in harmony with the facts ; and it was not 
 
216 LAW AND THEORY IN CHEMISTRY 
 
 long before Horstmann showed that the principles and 
 formula of thermodynamics generally are applicable 
 to all cases of chemical equilibrium. In other words, 
 the relations of thermodynamics and chemistry are not 
 limited to the application of one special thermodyna- 
 mical formula to a certain circumscribed class of equi- 
 libria distinguished as dissociations (see note 2, p. 199). 
 
 The verdict passed on Berthelot's " law " by thermo- 
 dynamics, is that only at absolute zero would such a 
 law obtain that only at the unattainably low tempera- 
 ture of 273 C. could prophecies founded on such a 
 law be relied on. For, presupposing the possibility 
 of chemical change at this temperature, it appears 
 that at absolute zero all reactions would be com- 
 plete ; all reactions would take place with an evolution 
 of heat, and of two or more possible reactions that 
 one would occur which is attended with the greatest 
 heat evolution. The higher the temperature rises above 
 the absolute zero, however, the more does Berthelot's 
 law "deviate from the truth inclining towards it." 
 With the rise of temperature above zero enters, 
 according to thermodynamic deduction, the possibility 
 of incomplete reactions of mobile equilibria, and, 
 moreover, the higher the temperature the more does 
 a given equilibrium shift in favour of the endothermic 
 partial change. 
 
 Thus, the displacement of HN0 3 from !JaN0 3 Aq 
 by H 2 S0 4 Aq is endothermic; the displacement of 
 H 2 S0 4 from Na 2 S0 4 Aq by HN0 3 Aq is exothermic. 
 Hence it follows that a rise in temperature shifts 
 
CHEMICAL EQUILIBRIUM 217 
 
 the equilibrium of the system NaN0 3 Aq, H 2 S0 4 Aq, 
 HN0 3 Aq, Na 9 S0 4 Aq in favour of the darker arrow 
 as shown in the following symbolism. 
 
 6 Aq + H 2 S0 4 Aq 
 Or, writing the reaction in equational form, 
 
 Na 2 N 2 6 Aq + H 2 S0 4 Aq = 
 
 (1 - x) Na 2 N 2 6 Aq + (1 - aj) H 2 S0 4 Aq, 
 
 we may state the same relation in another way by 
 saying that x decreases in value as the temperature 
 rises. 
 
 In cases where equilibrium is attained without 
 thermal change, temperature is without influence 
 thereon. Thus, nitric acid displaces hydrochloric 
 acid from salt solution without evolution or absorp- 
 tion of heat; the same is true for the displacement 
 of nitric acid from sodium nitrate solution by hydro- 
 chloric acid, hence in the equation 
 
 NaN0 3 Aq + HClAq = zNaClAq + a;HN0 8 Aq + 
 (1 - a) NaNOgAq + (1 - x) HClAq 
 
 the value of x is independent of the temperature. 
 
 The equilibria resulting from the etherification of 
 alcohols by acids (see p. 202) are also attained with- 
 out thermal disturbance, and are accordingly found 
 to be independent of temperature changes. 
 
 Having once and for all laid low the spectre of the 
 " Law of maximum work," thermodynamics replaced 
 
2i8 LAW AND THEORY IN CHEMISTRY 
 
 it with a general law which, though of somewhat 
 the same form as Berthelot's, differs from it in 
 being universally true at all temperatures. This law, 
 called the law of entropy, may be looked on as a 
 particular form of statement of the Protean second 
 law of thermodynamics. It states that a system is 
 in stable equilibrium only when its entropy is a maxi- 
 mum, and therefore that any change which entails an 
 increase in the entropy of the system is not only capable 
 of spontaneous occurrence, but will in fact actually 
 proceed until the entropy of the system reaches the 
 maximum value attainable under the conditions. 
 
 Unfortunately this quantity or function entropy is 
 a very difficult one to conceive, let alone measure, and 
 the law concerning it is to chemists of more theo- 
 retical interest than practical use. We can, however, 
 arrive at some kind of a conception of entropy by 
 the following considerations. 
 
 The electrical energy of an isolated charged body 
 is equal to the product JQV, where Q stands for the 
 quantity of electricity with which the body is charged, 
 and V is its potential. Suppose that for one body 
 the electric energy is JQ'V, and for another body 
 JQ"V". What are the conditions that electric energy 
 shall pass from one body to the other? Simply 
 electric connection and the inequality of V and V". 
 If V > V, then electric energy will pass from the 
 first body to the second until the potentials of both 
 bodies is the same, and this will happen even if the 
 electric energy of the first body is greater in amount 
 
CHEMICAL EQUILIBRIUM 219 
 
 than that of the second body. The passage of electric 
 energy from one system to another is independent of 
 the quantities of energy possessed by the two systems, 
 and depends solely on their potentials. Electric energy 
 always flows from places of high, to places of low 
 potential. 
 
 On this account V is regarded as the "intensity 
 factor " of electrical energy, and Q as the " capacity 
 factor." Energy of other forms can similarly be re- 
 solved into capacity and intensity factors. For instance, 
 the intensity factor of kinetic energy is velocity, 1 
 the capacity factor is mass, and so on. 
 
 What then are the factors of thermal energy ? The 
 intensity factor is very familiar to us and easily 
 capable of evaluation it is temperature. The capacity 
 factor does not appeal directly to our senses, nor can 
 we easily form a clear conception of it. Yet this 
 factor, which we are content to define rather than 
 conceive, has received the name entropy. Hence we 
 can restate the law of entropy in the following way : 
 Any change which can increase the value of the 
 capacity factor of the heat energy of a system takes 
 place with readiness, and the system only then attains 
 a position of stability (of disinclination to undergo any 
 further change) when the value of this capacity factor 
 is at a maximum. 
 
 In some forms of energy it is not possible to alter 
 the capacity factor by adding energy to the system. 
 
 1 Strictly speaking (velocity) 2 , the kinetic energy of a mass m moving 
 with velocity v being | wv 2 . 
 
220 LAW AND THEORY IN CHEMISTRY 
 
 The addition of kinetic energy to a moving bullet, for 
 instance, cannot alter the mass of the bullet, which here 
 represents the capacity factor. But in the case of heat 
 energy it is otherwise ; here it is possible, by adding 
 thermal energy to a system, to alter thereby the thermal 
 capacity factor of that system. 
 
 Suppose we have a mass of a solid substance S at 
 the absolute temperature T. Let us impart to S an 
 additional quantity of heat energy q under conditions 
 such that the intensity factor of the heat energy of the 
 system is not thereby altered, i.e., such that T remains 
 constant. This may be simply realised by supposing T 
 to denote the melting point of S, and q to be the latent 
 heat of fusion of S. Let us call the first state of the 
 system A and the final state B. To bring the system 
 to the state A from absolute zero requires an amount 
 of heat energy, say Q ; to bring the system from zero to 
 the state B requires Q + q units of heat energy. There- 
 fore in the passage from A to B the capacity of the system 
 for heat energy has evidently increased, and the increase 
 of heat capacity under the conditions named is called an 
 
 increase of entropy, and is measured by the quotient ? . 
 
 But thermodynamics has furnished us with tests of 
 stability involving functions other than entropy ; and 
 some of these tests possess the advantage of being more 
 easily applied to actual cases than the entropy test. 1 
 
 1 For details the reader is referred to Parker's Elementary Thermo- 
 dynamics, also to Ostwald's LehrbucJi der aUgemeinen Chemie, to which 
 the author is indebted for the treatment of the subject of entropy given 
 in the text. 
 
CHEMICAL EQUILIBRIUM 221 
 
 A momentary glance at the expression defining the 
 function known as the free energy, or the thermo- 
 dynamic potential at constant volume, throws clear 
 light on the question of Berthelot's "law." In the 
 equation 
 
 PA = UA - T A S A 
 
 F A denotes the free energy and U A the total energy of 
 a system in the state A. Its temperature in this state 
 is T A and its entropy S A . 
 
 Let us consider another state, B, of this system, for 
 which the values are F B , U B , T A , and S B , the volume 
 and temperature of the systems remaining the same in 
 both states. 
 
 From the two equations 
 
 FA = U A - T A S A 
 and 
 
 F B = U B - T A S B 
 
 we get by subtraction 
 
 (F A - F B ) - (U A - U B ) - [T A (S A - SB)]. 
 
 Now, it results from thermodynamic reasoning that 
 the change from the state A to the state B can only 
 proceed of itself when the free energy of the system is 
 decreased by the passage, i.e., (F A F B ) must be a 
 positive quantity. 
 
 Rearranging the equation we obtain 
 
 (U A - U B ) = positive quantity + T A (S A - S B ). 
 
 The value of S B being greater than that of S A , the 
 whole term T A (S A S B ) must be negative. 
 
 Of THE 
 
 UNIVERSITY 
 
222 LAW AND THEORY IN CHEMISTEY 
 
 Suppose the absolute magnitude of this term to be 
 less than the " positive quantity " term, then (U A U B ) 
 must obviously have a positive value ; that is to say, 
 the change of the system from the state A to the state 
 B has been attended by a loss of energy to the system 
 has been an exothermic change. But this is not the 
 only possible case. If the absolute magnitude of the 
 term T A (S A S B ) be greater than that of the positive 
 quantity term, then (U A U B ) must necessarily have a 
 negative value ; that is to say, the total energy of the 
 system has been increased by its spontaneous passage 
 from the state A to the state B the change has been 
 an endothermic one. 
 
 It is clear that the importance of the term T A (S A S B ) 
 rises pari passu with the temperature at which the change 
 proceeds. In other words, the higher the temperature 
 the more likely are endothermic reactions to occur. 
 
 THE END. 
 
 PRINTED BY BALLANTYNE, HANSON AND CO. 
 EDINBURGH AND LONDON. 
 
 tfttl 
 
14 DAY USE 
 
 RETURN TO DESK FROM WHICH BORROWED 
 
 LOAN DEPT. 
 
 This book is due on the last date stamped below, or 
 
 on the date to which renewed. 
 Renewed books are subject to immediate recall. 
 
 Iflan'WM 
 
 
 BEG'3 LD 
 
 
 jAiJG 1300 
 
 
 
 
 
 
 " ^* R VW 
 
 
 RFC'D LD 
 
 
 J/lN 1 n 'Cji -t nu 
 
 
 1 A O D*t * r M 
 
 
 
 
 FEB 16 198866! 
 
 
 REC'P 1 H 
 
 
 2'66-PAftL 
 
 
 1AY1419721? 
 
 
 MAY 
 
 (A1724slO)476B 
 
 Berkeley 
 
UNIVERSITY OF CALIFORNIA LIBRARY