Columbia Unibergttp 
 mtfteCitpofjSetogorfe 
 
 COLLEGE OF PHYSICIANS 
 AND SURGEONS 
 
 Reference Library 
 
 Given by 
 
Digitized by the Internet Archive 
 
 in 2011 with funding from 
 
 Open Knowledge Commons 
 
 http://www.archive.org/details/inauguraldissertOOgibb 
 
AN 
 
 INAUGURAL DISSERTATION 
 
 ON 
 
 CHEMICAL CLASSIFICATION 
 
 BY 
 
 OLIVER WOLCOTT GIBBS, A.M. 
 
AN 
 
 INAUGURAL DISSERTATION 
 
 ON A 
 
 NATURAL SYSTEM OF CHEMICAL CLASSIFICATION. 
 
 SUBMITTED TO THE PUBLIC EXAMINATION OF THE 
 
 TRUSTEES AND OF THE FACULTY OF MEDICINE 
 OF THE COLLEGE OF PHYSICIANS AND SURGEONS 
 
 IN THE CITY OF NEW YORK, 
 
 Under the authority of the Regents of the University of the State of New York, 
 ALEXANDER H. STEVENS, M.D., 
 
 PRESIDENT OF THE COLLEGE AND OF THE BOARD OF TRUSTEES, 
 
 FOR THE DEGREE OF DOCTOR OF MEDICINE, 
 On the 6th day of February, 1845. 
 
 By OLIVER WOLCOTT GIBBS, A.M. 
 
 OF THE CITY OF NEW YORK. 
 
 PRINCETON, N . J . : 
 
 PRINTED BY JOHN T. ROBINSON. 
 1845. 
 
fit- &f 
 
TO 
 
 JOHN TORREY, M,B, 
 
 PROFESSOR OF CHEMISTRY AND BOTANY 
 
 IN THE COLLEGE OF PHYSICIANS AND SURGEONS. 
 
 This Dissertation is Respectfully Inscribed. 
 
Natura non facit saitum. 
 
 Linnceut, 
 
INTRODUCTION. 
 
 The researches of chemists have resolved all known forms 
 of matter into fifty-six separate and distinct, species, which, 
 having thus far resisted every attempt at further decomposi- 
 tion are termed elements, and must in the present state of our 
 knowledge be regarded as simple substances. These Ele- 
 ments united in various proportions and with various degrees 
 of energy, constituted in the beginning the entire mass of the 
 Earth, and presented an Inorganic Nature, made up of vast 
 bodies of gaseous and liquid matter, as well as of many suc- 
 cessive series of rock, and of innumerable minerals, amorphous, 
 or bounded by harmonious and symmetrically disposed planes, 
 and constituting those regular figures which we term crystals. 
 At a later period in the genesis of Nature, the mysterious 
 power of Life was called into existence, and beginning per- 
 chance with the first cell-germ of the first lichen,- spread at 
 length over the whole Earth. Then innumerable new Forms 
 came into being/ and every plant, itself built up particle by 
 particle and cell by cell of inorganic matter, became a labora- 
 tory wherein chemical decompositions and recompositions 
 were perpetually going on through every step of growth, ma- 
 turity and decay. At a still late rperiod Life, gathering as it 
 were new energy with the lapse of time, passed from the 
 highest form of vegetable to the lowest form of animal organ- 
 izations, and thence upward in the scale of being through 
 every varying type and form, till it found in the structure of 
 man the limit beyond which it was not destined to pass. The 
 new and more complex organisms thus created, formed out 
 
8 CHEMICAL CLASSIFICATION. 
 
 of the same elements new and more complex chemical pro- 
 ducts, adapted to their growth, their structure and their wants. 
 The animal organism found in that of the plant, nearly all 
 the materials necessary for its own sustenance : of some of 
 these it changed the form without altering the chemical com- 
 position, while of others again it moulded the composition as 
 well as the form anew. And when at length Life ceased long- 
 er to exert over each individual its plastic force, the elements 
 obeying their mutual affinities formed new and simpler pro- 
 ducts, again to be brought under the sway of the vital force, 
 and thus destined to pass through countless cycles of change, 
 and to appear in the development and decay of myriads of 
 organized forms. Could an atom of Carbon or of Hydrogen 
 but speak to us what a history might it not unfold ! Through 
 every change however of the mineral, the plant, and the ani- 
 mal, the one universal law of Chemical combination — the law 
 of definite proportions— has from the beginning of things, 
 held good. Thus then we are led to recognize two great de- 
 partments of Chemical Science, the Chemistry of Unorganized 
 and the Chemistry of Organized bodies. The former embraces 
 all those combinations which take place under the operation 
 of purely chemical forces, whether in the laboratory of nature 
 or under circumstances determined by the will of man. The 
 latter comprehends all the products of the action of the vital 
 force upon inorganic matter, whether this action be immediate 
 as in plants or mediate as in animals, together with those 
 bodies which result from the former by the action of chemical 
 reagents. It is with the first of these departments that we 
 have at present to deal. Between the fifty-six elements into 
 which, as already stated, all known forms of matter have been 
 resolved, there exist numerous remarkable analogies, which 
 separate them first into subordinate natural families or groups, 
 and then by general though not indistinct resemblances, unite 
 them into one indissoluble chain, each link of which differs 
 rather in degree than in kind from its fellow on either side ; so 
 that the whote illustrates in unorganized nature, the truth of 
 
INTRODUCTION 
 
 the maxim of Linnaeus — Natura non facit saltum, nature makes 
 no leaps. And this then we assume as the fundamental idea 
 and central point of our Essay, namely, that this law of grades 
 which Linnaeus announced for the organized kingdom alone, 
 is an universal law, and prevails as well among lifeless atoms 
 as among living beings, and in the simplest crystals as well as 
 in the innumerable complex forms in which Life outwardly 
 manifests itself. The forms of nature are many, but Nature 
 herself is one. We propose then in the following pages to 
 trace these analogies as they exist between those forms of 
 matter which come within the range of Inorganic Chemistry? 
 whether these be simple or compound, and to point out the 
 most remarkable groups which such analogies constitute. 
 The time and the occasion will not permit us to do more, 
 though we would gladly extend our survey to the fertile and 
 as yet but little tilled field of Organic Chemistry. Moreover, 
 the classification of organic products must obviously depend 
 upon a prior classification of the elements themselves, and 
 upon an accurate knowledge of the nature and characters of 
 those compounds which are united in virtue of chemical 
 forces alone. By a chemical compound we shall of course un- 
 derstand any substance whose constituents are united in obe- 
 dience to the laws of definite proportions, and we shall gene- 
 rally speaking divide compounds into binary, ternary, quater- 
 nary, &c, according to the number of their ultimate or of their 
 proximate constituents. We may for the sake of convenience 
 consider the subject under three heads : 
 
 1st. The analogies between simple substances and their 
 division into groups. 
 
 2d. The analogies between compounds and their division 
 into groups. 
 
 3d. The analogies between simple and compound bodies. 
 
 Before taking up the special consideration of the several 
 groups, it will be well to dwell for a time in a general and 
 concise manner upon some of the more important points in 
 
10 CHEMICAL CLASSIFICATION. 
 
 which the analogies about to be pointed out may subsist, and 
 first then of the equivalent weights or combining masses. 
 
 1. The equivalents of a number of the elements appear to 
 be connected together by simple numerical relations. The 
 list of such instances of Isomerism originally given by 
 Dumas, has since been considerably extended by Dr. Kane, 
 and future researches, by correcting several of the equivalents 
 at present received, will probably exhibit other cases of coin- 
 cidence in this particular. Even now as we shall presently 
 show, we may without straining our figures, add several to 
 the list of instances enumerated by the last mentioned chemists. 
 The bodies between which such isomerisms exist, for the most 
 part resemble each other strongly in their chemical properties j 
 and general relations, notwithstanding the assertion of Dr. 
 Kane that " in no case are their properties more different." 
 The equivalent, numbers of very many compounds, however, 
 exhibit similar coincidences, without, at the same time, accom- 
 panying any analogies in composition or in chemical character. 
 Such instances must of course be regarded as accidental, and 
 must not be suffered in any degree to influence our systems of 
 arrangement and classification. We arrive then necessarily 
 at the conclusion that the chemical properties and relations of 
 bodies do not depend upon their equivalent weights alone, but 
 that in addition there are other and specific forces or properties 
 which are inseparably connected with the constitution of each, 
 and which bestow upon each its distinctive chemical char- 
 acter. 
 
 2. Admitting that the specific gravities of bodies represent 
 the relative weights of equal bulks, it follows that if we divide 
 the specific gravities by the atomic weights or equivalent num- 
 bers, we obtain the relative numbers of equivalents which 
 different substances contain under the same bulk or volume. 
 The numbers thus obtained have been termed by the German 
 chemist by whom they were first observed, the Atomic Num- 
 bers, and appear to be deserving of more attention than they 
 have as yet received. The atomic numbers of the elements 
 
 c 
 
ATOMIC NUMBERS. It 
 
 appear to be connected together in many instances by simple 
 ratios, and the simple substances constituting a natural group, 
 have usually the same number of equivalents contained under 
 the same volume. The requisite data for instituting a com- 
 parison between the atomic numbers of compound bodies are 
 for the most part wanting, since the specific gravities of very 
 few of these last have been accurately determined. It is very 
 much to be regretted, that this should be the case, since the 
 greater the number of equivalents which a given compound 
 contains under a given volume, the greater must be the energy 
 with which its constituents are united and vice versa, so that 
 we are thus apparently furnished with an accurate method of 
 determining Degrees of Affinity. To illustrate this by an ex- 
 ample, let us take the case of Hydrate of Potassa and of 
 monobasic sulphate of water. In a unit of volume the first 
 of these bodies contains 3032, the last 3766, equivalents. The 
 formula of the hydrate of Potassa is KO-f HO, that of the sul- 
 phate of water SO 3 -f- HO. Consequently the affinity of water 
 for sulphuric acid is greater than the affinity of water for 
 potassa, in the proportion of 3766 to 3032, or the ratio of its 
 affinity, for the two substances may be expressed by the frac- 
 tion 1.2420. In like manner we find that the affinity of sul- 
 phuric acid for water is greater than its affinity for potassa, in 
 the proportion of 3766 to 2745, so that if the principle which 
 we have laid down be correct, there must be something more 
 than a mere play of affinities exerted in the decomposition of 
 sulphate water, S0 3 +HO, by hydrate of potassa, KO + HO. 
 As a general rule, the greater the number of constituents which 
 any substance contains, the less is its atomic number, and it is 
 certain that a compound of two or more other compounds, 
 manifests for the most part affinities less energetic than those 
 of either of its constituents, and that in general the degree of 
 neutrality is directly proportioned to the complexity of the 
 composition. We shall endeavour to illustrate this position 
 more fully hereafter, when treating of the several groups. It 
 will be sufficient for the present to state the formula and the 
 
12 CHEMICAL CLASSIFICATION. 
 
 atomic numbers of three definite sulphates of water, and of 
 anhydrous sulphuric acid ; they are as follows : 
 
 S0 3 — 4910 
 
 HO + SO3— 3766 
 
 2HO + S0 3 — 3062 
 
 SHO-fSOg— 2431 
 
 No connection has as yet been traced between the atomic 
 numbers of a compound and those of its constituents. A con- 
 nection however of some kind or other doubtless exists, and 
 its discovery would be of the highest interest in a theoretic 
 point of view. 
 
 3. Intimately connected with the atomic numbers, are the 
 Combining Volumes, or the relative bulks of the equivalents 
 in the gaseous state. If we consider the space occupied by an 
 equivalent of Oxygen as unity, then an equivalent of Chlorine, 
 of Iodine, of Bromine, of Hydrogen, and of a number of other 
 bodies both simple and compound, occupies a space exactly 
 twice as great. The equivalent of sulphur on the other hand, 
 occupies, in the state of gas, but one-third the space which is 
 filled by an equivalent of oxygen, of selenium, and of tellu- 
 rium. It follows therefore that in one unit of space, there 
 may exist 3 equivalents of sulphur, i an equivalent of chlo- 
 rine or of hydrogen, or one equivalent of oxygen, selenium or 
 tellurium. To adopt for a moment the language of the ato- 
 mic theory, the atom of chlorine in the gaseous state is twice 
 as large as the atom of oxygen, the atom of sulphur but one- 
 third as large, and so on for all other bodies simple or com- 
 pound. The atomic volumes of all bodies then, while in the 
 gaseous state, are connected by simple numerical ratios, and 
 this fact is one of the deepest interest and importance. Of the 
 56 elements which are known to chemists, but a small propor- 
 tion have been obtained or are found in the gaseous state. 
 The combining volumes of a great number cannot therefore 
 be experimentally determined, though in many cases they 
 have been deduced a priori from clear and distinct analogies. 
 The identity of two or more substances in this particular. 
 
ISOMORPHISM. 13 
 
 does not necessarily involve a similarity in general chemical 
 properties and relations as will hereafter be ^hown. 
 
 4. By a process exactly the reverse of that employed in 
 obtaining the Atomic Numbers, namely, by dividing the equi- 
 valent weights by the specific gravities, Dr. Kopp has ob- 
 tained a series of numbers which he regards as expressing the 
 atomic volumes of bodies; or in other words the relative 
 volumes of their atoms ; and he considers that his researches 
 justify the conclusion that isomorphous bodies have the same 
 atomic volume. The apparent deviations from this law ex- 
 hibited by many isomorphous bodies, he regards as aris- 
 ing from the fact that substances are not absolutely iso- 
 morphous but only approximately so, and a glance at the 
 numbers which he brings forward in support of his opinion, 
 will abundantly justify the necessity of admitting this or some 
 similar explanation. 
 
 5. Almost every unorganized body may be obtained in a 
 crystalline state, and may consequently be arranged under 
 some one crystallographic system, and have its interfacial and 
 interaxial angles determined with a greater or less degree of 
 accuracy. If now we institute a comparison between the 
 crystalline forms of different substances simple or compound, 
 we find that between those which are connected together by 
 strong chemical analogies, there exists also certain definite re- 
 lations of form, in many instances approximating toward iden- 
 tity. A natural family or group has in general the same crys- 
 talline as well as the same chemical types. Identity of crys- 
 talline form however by no means necessarily implies analogy 
 in chemical constitution, since a substance composed of a great 
 number of molecules of different kinds may crystallize in the 
 same form as an element, and since on the other hand two 
 elements may appear in crystallographic systems which have 
 no connexion with each other. In fine, so numerous are the 
 instances of agreement in chemical constitution and relations 
 without, a corresponding analogy of form, and again of a coin- 
 cidence in crystalline form without a corresponding agree- 
 
 2 
 
14 CHEMICAL CLASSIFICATION. 
 
 ment in chemical relations, that while similarity of form must 
 ever be regarded as a strong reason for classifying together 
 bodies which are otherwise analogous, yet the want of such a 
 similarity will never of itself justify the separation of substances 
 which agree in their general chemical relations. It is the too 
 implicit reliance upon analogy of form which renders the 
 classification proposed by Graham imperfect, and which has 
 led him to associate together substances which differ widely 
 in their chemical relations, and to separate others between 
 which there exists in these respects the strongest resemblance. 
 Certain bodies are met with crystallized in two different forms 
 and isomorphous with two distinct classes or groups of other 
 substances. Such bodies are said to be dimorphous, and serve 
 to connect several groups together at once. The relatior^ of 
 dimorphism is subject to the same restrictions as that of iso- 
 morphism and must not be too strongly relied upon in classi- 
 fying and arranging substances in natural families. No con- 
 nection has yet been traced between the crystalline form of a 
 compound and the crystalline forms of its constituents. The 
 problem is one of the finest which offers itself to the enquiring 
 mind, in the whole range of physical science, but sufficient 
 data to form the basis of the solution are yet wanting. The 
 chemical and physical history of no compound can be consid- 
 ered complete without an accurate and scrupulous determina- 
 tion of its crystalline form, and the negligence of chemists with 
 regard to this most important particular cannot be too strongly 
 reprobated. We shall hereafter point out a few instances in 
 which a direct relation subsists between the forms of a com- 
 pound and the forms of its elements. It may not be improper 
 to note in this place the fact that as a general rule, the simpler 
 the constitution of a body, the simpler also is its crystalline 
 form and vice versa. 
 
 6. The relations of bodies to heat, though not sufficient m 
 themselves to furnish grounds for associating them in groups 
 are yet of very great interest and importance. The researches 
 of Begnault and of Newmann and Awgadro, have recently 
 
ISOMORPHISMS. 15 
 
 afforded strong confirmation of the accuracy of Dulong and 
 Petit's law, that the specific heats of equivalent weights of sim- 
 ple substances are equal. The few cases which appear to be 
 exceptions to this law may easily be reconciled with it, not by 
 an arbitrary subdivision of the atomic weights, but by a differ- 
 ent enunciation of the law itself. Thus if we state that " the 
 specific heats of the elements are either the same or are con- 
 nected with each other by simple multiples," we obtain a clear 
 and satisfactory explanation of the discrepancies which have 
 thrown so much doubt upon the existence of any definite law 
 whatever, and which have led to so many alterations of the 
 equivalents and to so many improbable hypotheses. Between 
 the specific heats of compound bodies there exist, also, sim= 
 ilar relations, though they have not as yet been very fully 
 studied. No connexion has been detected between the fusing or 
 boiling points of different substances whether simple or com- 
 pound and their chemical constitution or relations ; though Ger- 
 hardt has recently succeeded in deducing an empirical formula, 
 by which, having given the constitution and the boiling point of 
 one liquid hydrocarbon and the constitution of a second, he is 
 able to determine the boiling point of this last, by calculation 
 alone. No connexion appears to subsist between the fusing 
 or boiling point of any substance, and the number of equiva- 
 lents which it contains in the unit of space. The quantity of 
 heat which is developed during the act of chemical combina- 
 tion appears always to be definite, and depends upon the na- 
 ture and composition of the substances uniting. The subject 
 however has not yet been sufficiently investigated to admit of 
 the deduction of definite laws. The same remark may be 
 made with regard to the latent heats of all substances what- 
 ever. We are as yet very imperfectly acquainted with the 
 subject. Neither the conducting powers nor the expansibili- 
 ties of bodies appear to be connected with their chemical con- 
 stitution. 
 
 7. Upon the consideration of the electrical relations we do 
 not propose in this place to enter. The subject is as yet too 
 
16 CHEMICAL CLASSIFICATION. 
 
 little understood to admit of anything more than a general ap- 
 plication of these relations to the purpose of classification. 
 We shall point out in speaking of the several groups, the elec- 
 trical conditions of each, agreeably to the theory of Faraday 
 and Berzelius, though we do not propose to adopt this as the 
 basis of classification, as is now very commonly done. 
 
 8. The influence which the chemical constitution of a body 
 exerts upon the transmission, reflection, absorption or polari- 
 zation of Light, has hitherto excited very little attention. No 
 connexion has been traced between the refractive power of a 
 compound and the refractive powers of its constituents, and 
 very few data exist with which to institute a comparison- 
 Possibly the action of bodies upon Light may be] connected 
 with their Atomic Numbers, but the whole subject is involved 
 in the deepest obscurity and presents a noble field for experi- 
 mental research. 
 
 9. It remains to consider in a brief and general manner, the 
 purely chemical relations of bodies to each other, that is to say, 
 the relations of constitution, of acidity and basidity, of molecu- 
 lar types and of general character and properties. The mole- 
 cular type of any substance is of the utmost importance in the 
 determination of the group to which it belongs. As the ele- 
 ments are to the best of our knowledge identical in this partic- 
 ular, we are obliged, in order to investigate their chemical re- 
 lations, to examine the molecular structure of the compounds 
 which they form with one another, and from these to deduce 
 the general character of each. We find then that a natural 
 group of Elements, by entering into combination with other 
 elements, forms a series ot compounds whose molecular struc- 
 tures are precisely similar to one another, and which may be 
 transformed the one into the other by a simple change in the 
 chemical symbols which constitute the formulae. In like man- 
 ner a natural group of compound bodies is composed of sub- 
 stances which are exactly parallel to one another, both in com- 
 position and in the arrangement of their constituents. Bodies 
 which are similarly constituted exhibit for the most part very 
 
GROUP I. 17 
 
 similar relations toward other bodies, and conversely we are 
 often able to decide upon the constitution of a body by the 
 analogy which exists between its properties and those of an- 
 other substance whose composition is already known. The 
 relations of acidity and basidity depend strictly upon molecu- 
 lar type, and are purely relative, a substance which in one 
 connexion is an acid, is in another a base, and in another 
 again, neutral. The neutrality of any compound is for the 
 most part in proportion to the number of secondary com- 
 pounds of which it is composed, neutrality being here used in 
 the sense of incapacity to enter still further into combination. 
 Certain of the elements generate by uniting with other ele- 
 ments compounds, which are either acids or bases; others 
 again generate new substances which are eminently neutral 
 and incapable of forming new compounds with one another. 
 The union of certain bodies is accompanied with the develop- 
 ment of light and heat, with others again no such phenomena 
 are visible during the act of combination. Many substances 
 are capable of uniting directly with each other, when brought 
 together in a free state and at ordinary temperatures, the affin- 
 ities of others again must be modified by heat, light or elec- 
 tricity, before they can be made to combine. Finally the 
 three states in which bodies present themselves, as solids, 
 liquids or gases, appear to depend upon their relations to heat 
 alone, and to have no definite relations to their chemical con- 
 stitution or properties. 
 
 We have thus taken a brief survey of some of the most 
 important points in which the chemical analogies of bodies 
 subsist, and we now proceed to the consideration of the sev- 
 eral groups which are constituted by the elementary substan- 
 ces. 
 
 GROUP I. 
 
 The first group comprises eight elements. Oxygen, Sul- 
 phur, Selenium, Tellurium, Chlorine, Bromine, Fluorine. Be- 
 tween these eight elements, though widely different in the 
 
18 CHEMICAL CLASSIFICATION. 
 
 common physical properties of matter there exist the most 
 intimate chemical relations, so that as a whole the group con- 
 stituted by them is more strongly marked than any other. 
 These relations are as follows : 
 
 1. A body belonging to any other group commonly unites 
 with a member of this, with that evolution of light and heat 
 which chemists term combustion. 
 
 2. The affinities of the members of this family are more 
 energetic and are exerted through a wider range than those 
 of any other family. And this is proved by the number and 
 variety of the compounds which they form with other sub- 
 stances ; by their combining, as a general rule, at much lower 
 temperatures than other substances, and by their capacity of 
 entering into direct combination, while in a free state. 
 
 3. The eight elements composing this group are capable of 
 replacing each other in a compound of given molecular type 
 without altering that type, and without altering, otherwise 
 than in degree, the chemical relations of the primitive whence 
 the new substances are by substitution derived. The new 
 substances will be acids, bases or neutral, according as that 
 from which they are supposed to be derived was acid, base, 
 or neutral. 
 
 4. The existence of any particular compound or compounds 
 of a substance not belonging to group 1st, with any member 
 of this group, is an a priori evidence, that there may also ex- 
 ist similar compounds of the other members of group 1st with 
 the substance in question. Thus the different groups of chlo- 
 rides, oxides, sulphides, &c. are exactly parallel to one another 
 in composition, and as already remarked, (3) differ in their 
 chemical relations only in degree. 
 
 5. A compound of one equivalent of a member of this 
 family, with one equivalent of a radical simple or compound, 
 performs in general the part of a base, though as already re- 
 marked one base may act the part of an acid toward another. 
 As the number of equivalents of the substance belonging to 
 group 1st, increases, the acidity of the compound also in- 
 
GROUP I. 19 
 
 creases. A binoxide or bichloride is feebly basic toward pow- 
 erful oxacids or chloracids, and feebly acid toward powerful 
 oxy-bases or chlorobases. A teroxide or terchloride is in almost 
 all cases a powerful acid, and a pentoxide or pentachloride 
 still more so. 
 
 6. Substitutions of one member of group first for another 
 may be either complete or partial. Thus chlorine may either 
 replace 3 equivalents of oxygen in a teroxide, or only one, 
 We meet for example with such compounds as S0 2 C1, Cr0 2 Cl, 
 Fe 2 2 Cl, COC1, and many others of a similar constitution. 
 Such compounds will of course be most conveniently classed 
 with the substance from which they are or may be considered 
 to be derived. 
 
 7. Between the equivalent weights of several members of 
 this group remarkable relations are found to subsist. Thus, 
 Oxygen, Sulphur, and Tellurium appear to be isomeric, their 
 equivalents being to each other as 8, 16 and 64, very nearly, or 
 as 1,2 and 3. Selenium on the other hand, which very closely 
 resembles Sulphur in its chemical relations, is entirely uncon- 
 nected with it in this particular, but its equivalent is very 
 
 nearly half that of Bromine. Se=39.63= /8,39 =39.19=^ Br. 
 
 2 
 
 Chlorine, Iodine and Bromine do not appear to be connected 
 by any definite relations between their atomic weights, and 
 future researches may show that there is no real isomerism be- 
 tween Selenium and Bromine.* 
 
 8. The combining volumes of Oxygen, Selenium and Tel* 
 lurium are equal to each other and are represented by unity. 
 The combining volume of Sulphur is 3 times, and that of 
 Chlorine, Iodine, Bromine and Fluorine twice, as great as that 
 of Oxygen. Hence either the atom of Chlorine is twice as 
 large as the atom of Oxygen or else Oxygen contains in the 
 same space twice as many particles or atoms as Chlorine does. 
 This last is perhaps the more probable opinion of the two. 
 
 * According to Dumas, the equivalent of Selenium is 40, which makes it iso- 
 meric with Oxygen: 8 X 5 = 40. 
 
20 CHEMICAL CLASSIFICATION. 
 
 9. The atomic numbers of Chlorine, Iodine and Bromine 
 are the same, that is to say, one unit of space contains an equal 
 number of equivalents of each of these substances. The 
 numbers themselves are, Chlorine 383, Bromine 380, Iodine 
 392, the mean of which is 385, which may therefore be con- 
 sidered as representing the true atomic number of the 3 sub- 
 stances in question. The atomic numbers of Sulphur and Se- 
 lenium are apparently unconnected ; they are, of Sulphur 125, 
 of Selenium 109. As Oxygen has not been obtained in a 
 solid, and Fluorine in an isolated state, their atomic numbers 
 cannot of course be determined. Tellurium is connected nei- 
 ther with Chlorine and Bromine, nor with Sulphur or Selenium 
 by any very certain relations of atomic numbers. It contains 
 in one unit of space 978 equivalents, which is very nearly 9 
 times as many as are contained within the same volume of Se- 
 lenium. 109x9 = 981 = 978 + 3.J 
 
 10. As a general rule an oxide appears to contain under 
 the same bulk a greater number of equivalents than a corres- 
 ponding sulphide, and a sulphide a greater number than a cor- 
 responding chloride. Such at least is the conclusion which 
 the writer has drawn from his own calculations which are all 
 which he has had to rely upon. The following tables will 
 exhibit this to a limited extent. (Note. The specific gravity 
 of water is here taken as 1000.) 
 
 Pb=11030, PbO = S504, PbS = 6336, PbCl~ 4080, Aq= 
 9640, AqO=6139, AqCl=3894. 
 
 Hg=13210,HgO= s 10119, HgCl = 3754,Hg 2 2 = 8097,Hg 
 2C1=2932. 
 
 Perhaps a more extended induction may overthrow the 
 opinion which we have advanced in regard to the atomic 
 numbers of oxides, sulphides and chlorides. But if on the 
 contrary more accurate and extensive observation shall de- 
 monstrate its correctness, it will then as it seems to us, afford 
 an explanation of many facts in the history of these bodies 
 which are at present involved in obscurity. For does not a 
 very great degree of condensation imply a very energetic at- 
 
GROUP I. 21 
 
 traction ? and does not a lesser degree of condensation imply 
 a weaker attraction ? The subject is at any rate worthy of a 
 most thorough and careful study. 
 
 11. Between the atomic volumes of certain members of 
 this group there also exist remarkable relations. Thus by di- 
 viding the atomic weights of Chlorine, Iodine and Bromine 
 by their respective specific gravities we obtain as quotients, 
 25, 25, 26, and we therefore conclude that the atomic volume 
 or the bulk of the equivalent is the same for each of these 
 bodies. Thus not only are there as we have already shown 
 (9) the same number of equivalents in the same bulk of each 
 of these bodies, but the bulk or volume of each of these equiv- 
 alents is also the same. And this last is by no means a ne- 
 cessary consequence, as might at first appear, of the first. For 
 as it is clear that matter, whether it consist of atoms or not, is 
 not continuous through space, so it must be evident that one 
 unit of space, one cubic inch for example, may contain either 
 very different numbers of equal-sized atoms or the same num- 
 ber of atoms differing in size, or again, different numbers of 
 different atoms, provided only that these atoms are unequally 
 distant from each other in the three cases which we have sup- 
 posed. The volume of the atom of Tellurium is to the volume 
 of the atom of Sulphur or Selenium as 10 to 8 or as 5 to 4, and 
 to the volume of the atom of Chlorine, Iodine or Bromine as 
 10 to 25 or 2 to 5. The atomic volumes of Oxygen and Flu- 
 orine as well as their atomic numbers are of course unknown. 
 
 12. Of the 8 elements which compose group 1st, Sulphur 
 and Selenium are isomorphous. Chlorine, Bromine and Io- 
 dine are also isomorphous, and possess besides the same atomic 
 volume and the same atomic number. The compounds of 
 Fluorine are also uniformly isomorphous with those of Chlo- 
 rine, Iodine and Bromine, and we have therefore strong rea- 
 sons for regarding Fluorine as itself isomorphous with Chlo- 
 rine, Iodine and Bromine. Tellurium is not connected with 
 the other members of Group 1st by isomorphism, unless in- 
 deed it should prove like Sulphur to be dimorphous. Finally, 
 
22 CHEMICAL CLASSIFICATION. 
 
 the crystalline forms of the sulphur salts have not been suffi- 
 ciently studied to enable us to form any idea of the crystalline 
 form of Oxygen, by instituting a comparison between the two 
 classes of salts, which Sulphur and Oxygen are respectively 
 capable of generating. We may notice in this connexion a 
 fact which does not seem to have attracted the attention of 
 chemists. The crystals of many anhydrous protochlorides, 
 protiodides, protobromides, and protofluorides belong to the 
 regular system, being for the most part either cubes or regu- 
 lar octahedra. Furthermore, the crystals of many double pro- 
 tochlorides, double protiodides, &c. are also met with in the 
 regular system. In order not to leave this assertion without 
 proof or illustration, we will cite the formulae of a few com- 
 pounds together with their crystalline forms. 
 
 (KC1, NaCl, AqCl, KI, KB^ CaF, NaF, NaB 3 ,) cubes. 
 (AqCl+KCl), (AqCl + NaCl, AqCl + HCl,) cubes. Manypro- 
 tochlorides and protobromides crystallize with water, and their 
 form is thus rendered more complex. Other relations of form 
 and composition will be pointed out when treating of other 
 groups. 
 
 13. The specific heats of certain of the elements connected 
 with this group appear to bear a definite relation to each other. 
 Thus the specific heat of an atom of Sulphur being consid- 
 ered as unity, that of an atom or equivalent of Tellurium is 
 represented by 2, or is twice as great, that of Iodine again is 
 4 times as great as that of Sulphur and twice as great as that 
 of Tellurium. The specific heats of equivalent weights of the 
 other members of group 1st have not yet been accurately de- 
 termined. 
 
 14. The elements composing this group are more highly 
 electro-negative than any other substances which we are ac- 
 quainted with. In the electro-chemical theory they constitute, 
 par excellence, the electro-negative group, and they are sepa- 
 rated from the other groups in this respect by a wide interval. 
 If a compound of an element belonging to this family with 
 one belonging to any other, be submitted to electrolysis, the 
 
GROUP I. 23 
 
 constituent belonging to the first group will always be found 
 at the positive pole, that belonging to any other group will on 
 the contrary always be found at the negative pole. An elec- 
 tro-negative element is termed by Graham, Chlorous, while on 
 the other hand an electro-positive element or constituent re- 
 ceives the appellation of Zincous. The different members of 
 the group differ in the degree of electro-negativeness, but our 
 knowledge of the subject is far from being sufficiently com- 
 plete to enable us to form a scale in which each shall find its 
 proper place, or even to say positively which of the elements 
 is most highly electro-negative, and which again is at the other 
 end of the scale. This and much else is open for investiga- 
 tion. 
 
 15. It remains to consider the different members of this 
 group in their relations with each other. With the exception 
 of Fluorine and of course Oxygen, they all agree in forming 
 with Oxygen, compounds which for the most part are power- 
 fully acid. The types of these compounds are not in the pre- 
 sent state of our knowledge perfectly parallel. Sulphur forms 
 with Oxygen 6 distinct acids, of which but one, namely, hypo- 
 sulphuric acid has its analogue among the Oxygen compounds 
 of Chlorine, Iodine and Bromine, while two, namely, sulphur- 
 ous and sulphuric acids, have analogues among the com- 
 pounds of Selenium and Tellurium. Chlorine again forms 4 
 compounds with Oxygen, to which Iodine offers but two par- 
 allels, and Bromine but one. Finally, Selenium offers one com- 
 pound, the protoxide, which agrees in composition with hypo- 
 delorous acid or protoxide of Chlorine. In order to exhibit 
 at once the general types of the compounds of this group* 
 we will assume the letter A to represent any member of the 
 group in a general formula, accenting it thus A' or A" when 
 necessary to consider or represent more than one element at 
 the same time and in the same formula. Adopting these sym- 
 bols then we may construct general formulae which will re- 
 present the types of all known combinations of the members 
 of this group with one another, and these types we find to be 
 as follows : 
 
^4 CHEMICAL CLASSIFICATION. 
 
 AA'j— AA } 2 — AA' 3 — AA' 4 — AA'_— AA' 7 — A 2 A' a - 
 
 A 2 A V— A 3 A ' 5 — A 4 A V 
 Corresponding to which we have 
 
 S> 
 
 AA' r 
 
 -C10- 
 
 -SeO- 
 
 — TeCl— 
 
 
 
 
 AA' 2 - 
 
 -S0 2 - 
 
 -SeO 
 
 2 -Te0 3 - 
 
 -TeCl 2 - 
 
 ~C1S 2 ~ 
 
 -SCi 2 
 
 AA 'a- 
 
 -so,- 
 
 -SeO 
 
 3 ~Te0 3 - 
 
 -S0 2 C1- 
 
 ^SOJ- 
 
 -S0 2 ! 
 
 AA' 4 - 
 
 -C10 4 . 
 
 
 
 
 
 
 AA' 5 - 
 
 -cio,. 
 
 10.. 
 
 B 2 5 -B 
 
 2 C1 6 -IC1 
 
 2 5. 
 
 AA' 7 - 
 
 -C10 7 - 
 
 -10, 
 
 , 
 
 
 
 
 A 2 A' 2 
 
 -s 2 o, 
 
 2 ' 
 
 
 
 
 
 A 2 A *5 
 
 — s„o 
 
 5 . or 
 
 so 2 +so 3 
 
 1 
 
 
 
 A 3 A '5' 
 
 -s,o, 
 
 j. or 
 
 S0 3 + S0 3 
 
 s? 
 
 
 
 A 4 A' 5 - 
 
 -S 4 ; 
 
 > * 
 
 
 
 
 
 We have already alluded to the number and characters oi 
 the compounds which the members of this group form with 
 those of other groups, and this portion of the subject will also 
 meet with further consideration in speaking of the other 
 groups, specially and individually. 
 
 The bodies which we have arranged together under this 
 first class compose the first and fifth classes of Graham. They 
 are collectively termed Basacigen by Dr. Hare, in reference 
 to their property of forming both acids and bases by uniting 
 with other substances. By the latter chemist also this group 
 is again, in deference to the opinions of Berzelius, subdivided 
 into two, the first termed Amphigen, comprising Oxygen, Sul- 
 phur, Selenium, Tellurium ; the second termed Halogen, com- 
 prising Chlorine, Iodine, Bromine and Fluorine. As the term 
 Amphigen has not come into very general use, it might be 
 well to substitute it for the term Basacigen, which though 
 very expressive in itself is not susceptible of etymological 
 changes. Should it be thought worth while to subdivide the 
 whole group into two sub-groups, then for the same reason 
 that Chlorine, &c. are termed Halogen, Oxygen, Sulphur, &c. 
 might be termed Alkaligen, as has indeed been proposed by 
 some German chemists. Perhaps the best nomenclature after 
 all would be the simplest, viz ; to associate all the elements 
 under the title of the Oxy-Chlorine Family, 
 
GROUP IT. 25 
 
 GROUP II. 
 
 The second group comprises Nitrogen, Phosphorus, Arsenic 
 and Antimony. The parallel between the chemical proper- 
 ties and relations of these four elements is for the most part 
 very striking. If we assume the letter R as a general symbol 
 for the whole group, while A represents, as before, any ele- 
 ment belonging to group 1st, we shall find the following types 
 for the compounds of the two groups. RA — RA 2 — RA 2 — 
 RA 4 — -RA 5? corresponding to which we have 
 
 to RA— NO— PO 
 
 to RA 2 — N0 2 — AsS 2 — 
 
 t0 RA 3 — NO3—NS3— PO.— PS 3 — PCI3— PBr 3 — PF 3 — 
 As0 3 — Sb0 3 — AsS 3 — SbS 3 — AsCl 3 — SbCl 3 +&c. 
 
 to RA 4 — N0 4 — Sb0 4 — SbS 4 — SbCl 4 . 
 
 to RA,— NO,— P0-— AsO,— Sb0 5 + &c. &c. 
 
 In addition to these, phosphorus forms a suboxide to which 
 there is no analogue among the other elements of the 2d 
 group. Besides the analogies of the compounds which the 
 members of this group form with those of group 1st, there 
 are also other and not less remarkable relations. Each mem- 
 ber of group 2d forms with hydrogen a gaseous compound 
 of the type RH 3 . These compounds are severally represented 
 by the formulas NH 3 , PH 3 , AsH 3 , SbH 3 . The first and se- 
 cond of these unite with the halides of hydrogen to form defi- 
 nite compounds, strongly analogous to metallic halides and of 
 the type RH 3 4-HA\ The first again, ammonia, is capable 
 also of uniting with the oxides, sulphides, &c. of hydrogen 
 to form powerful oxybases and sulphobases, which moreover 
 are isomorphic us with similarly constituted oxides and sul- 
 phides of potassium. To these compounds, the phosphuret, 
 arseniuret and antimoniuret of hydrogen offer no parallels. 
 Again, ammonia and phosphuretted hydrogen are capable of 
 forming a very numerous class of compounds by combining 
 directly with oxides, chlorides, &c. as well as with very many 
 classes of salts composed of acids and bases. Antimoniuret- 
 ted and arseniuretted hydrogen appear to form no such com- 
 
 3 
 
26 CHEMICAL CLASSIFICATION. 
 
 pounds. One equivalent of the hydrogen of ammonia may 
 be replaced by an equivalent of metal forming a body of the 
 type NH 2 M, and presenting characters very analogous to those 
 of a protoxide or protochloride. We do not know that either 
 of the compounds of the other three members of this group 
 with hydrogen are susceptible of a similar molecular replace- 
 ment. Lastly, by means of a voltaic current, certain of the 
 compounds of ammonia, those namely, with the halides of 
 hydrogen, may be decomposed, the ammonia uniting with an 
 equivalent more of hydrogen to form a compound metal which 
 forms with the mercury which constitutes the negative pole of 
 the battery, a true amalgam, while the halogen constituent is 
 given off at the positive pole of the circuit. We are as yet 
 ignorant whether the compounds of phosphuretted hydrogen 
 with the halides of hydrogen are or are not capable of under- 
 going a similar decomposition, and whether or not there may 
 be a compound of phosphorus and hydrogen exactly analo- 
 gous to the hypothetic ammonium, and like it forming an 
 amalgam with mercury. And it would be a most deeply in- 
 teresting experiment, to submit to electrolysis in contact with 
 mercury the compound of phosphuretted hydrogen and iodo- 
 hydric or bromohydric acid, represented by the formula 
 PH 3 -j-HI or PH 3 +HBr, and to determine whether these 
 compounds are not, like the iodide and bromide of ammo- 
 nium, in all chemical relations, true iodides and bromides and 
 not idohydrates and bromohydrates. It has been shown by 
 Rose that a compound of phosphorus, hydrogen and chlorine, 
 exactly analogous to sal-ammoniac, exists in combination with 
 bichloride of titanium, forming a lemon-yellow substance, 
 whose formula is PH 4 Cl-f-3TiCl 2 . The combinations of the 
 members of this group with hydrogen, illustrate the law of 
 grades in a very striking manner, those chemical properties 
 and relations which are most strongly marked in the nitrogen 
 compounds, shade off as it were first into the phosphurets, 
 then the arseniurets, and finally are lost, or at best very feebly 
 exhibited in the antimoniurets, of hydrogen. In the com- 
 
GROUP if, 27 
 
 pounds of the members of this family with those of the first 
 group, equally striking analogies are manifested. For exam- 
 ple, the nitric and antimonic acids N0 5 and Sb0 5 unite with 
 but one equivalent of base, phosphoric acid P0 5 unites with 
 one, two, or three, while arsenic acid again As0 5 , unites with 
 three equivalents of base alone. For the formulas of the tri- 
 basic salts of the two last mentioned acids, we have 
 PO 5 + Na02HO— PO 5 + 2NaO,HO— PO - + 3NaO, 
 AsO 5 -f Na0.2HO— AsO 5 + 2NaO.HO— AsO 5 + 3NaO, 
 which formulas it will be seen are mutually convertible, by 
 the simple interchange of the symbols for phosphorus and 
 arsenic. Though arsenic acid forms but one class of salts, 
 yet the type of the three classes of salts formed by phospho- 
 ric acid is not lost with the comp^ 11 ^/ 1 ' f nic, but dis- 
 tinctly manifests itself in th : we have 
 the formulas of the three cIe iz : the mo- 
 nobasic, the bibasic and the s follow 
 
 P0 5 +NaO— P0 5 
 corresponding to which we 
 
 AsS 5 +NaS— AsS 5 f- £NaS— AsS 6 4-3Na 
 formulas which are mutually convertible, by the interchange 
 of symbols for oxygen and sulphur. Here again the law of 
 grades is beautifully exhibited. We are as yet unacquainted 
 with any tribasic sulpharseniate, corresponding to the tribasic 
 arseniates and phosphates whose formulas are given above, 
 and which contain both a stable and an unstable base. It is 
 clear that should such compounds ever be formed, sulphides 
 of hydrogen will take the place in the sulpharseniates of the 
 oxide of hydrogen in the oxyarseniates, so that we should 
 have 
 
 AsS 5 + NaS.2HS— AsS 5 + 2NaS.HS— AsS 5 + 3NaS, 
 corresponding to 
 
 AsO 5 + Na0.2HO— AsO 5 + 2NaO.HO— AsO 5 + 3NaO. 
 The salts formed by the sulphides of nitrogen and phos- 
 phorus have not yet been studied and are well worthy of at- 
 tention. The nitric and antimonic acids as already observed 
 
28 CHEMICAL CLASSIFICATION. 
 
 unite with but one equivalent of base ; on the other hand 
 however it appears certain that the sulphantimonic acid is tri- 
 basic, the formulae of several of its salts being as follows: 
 
 SbS 5 + 3BaS + 6HO— SbS 5 ,3Nas+18HO— $bS 57 3AqS. 
 which obviously correspond with the tribasic phosphates, ar- 
 seniates and sulpharseniates. Still further, the nitrous and 
 antimonious acids N0 3 and Sb0 3 appear to unite with but 
 one equivalent of base like the nitric and antimonic acids. 
 Phosphorous and arsenious acids on the other hand are in all 
 probability tribasic, all the known salts of the former acid 
 being constituted upon ^)he formula P0 3 +3RO while those 
 of arsenious acid appear to come under the formulae As0 3 4- 
 3RO, As0 3 +2RO, As0 3 +RO, and thus to be monobasic 
 and bibasic as well as tribal:- In like manner the formulae 
 tlpfaarseni e an exact analogy between 
 
 i and the art; lae of the two classes of salts 
 
 As0 3 + (1,2. . > 3 + (l. 2. 3)RS. Sulphar- 
 
 senlous acid appears gous to sulphantimonious 
 
 acid as the two form i i with basic sulphides nu- 
 
 merous native salts or l is whose formulae are precisely 
 
 similar. The general formulae of the sulphantimonites like 
 those of the sulpharseniates may be expressed bySbS 3 -f 
 (1.2. 3)RS, corresponding to which we have 
 
 Zinkenite SbS 3 + PbS— Feather Ore SbS 3 +2PbS— Boul- 
 angerite SbS 3 + 3PS— Ruby Silver SbS 3 + 3PbS. 
 
 We have dwelt upon sulphur salts of the members of this 
 group at considerable length, both on account of their intrinsic 
 interest, and because, though their formulae are stated in many 
 works on Chemistry, no deductions whatever appear to have 
 been drawn from them. They appear to us to illustrate in a 
 very striking manner the law of grades, and to point out 
 many paths of research upon which at some future period we 
 hope to enter. 
 
 Of the 4 elements which compose this group, two, viz : 
 arsenic and antimony are isomorphous in their free or uncom- 
 bined state. The salts of phosphoric and arsenic acids which 
 
GROUP II. 29 
 
 are similar in constitution correspond also in form. No cer- 
 tain isomorphism connects the compounds of hydrogen with 
 this group, while on the other hand the isomorphism of anti- 
 mony and arsenic with tellurium preserves a resemblance in 
 this respect between groups first and second. 
 
 Notwithstanding the isomorphism of arsenic and anti- 
 mony, the numbers representing their atomic volumes do 
 not correspond, the number for antimony being according to 
 our own calculation 19.22 while that for arsenic is 12.68. 
 Hence there must be some exceptions to the law of Dr. Kopp. 
 There is a similar disagreement between the atomic numbers 
 of the two substances, the number for antimony being .0519 
 while that for arsenic is .0792.* Yet it seems to us that these 
 apparent discrepancies rightly considered, will entirely disap- 
 pear. Adopting for the sake of simplicity the atomic theory, 
 we reason as follows : arsenic and antimony are isomor- 
 phous — the volume of the atom of arsenic is less than that 
 of antimony in the proportion of 12.68 to 19.22, but in the 
 same bulk a crystal of arsenic contains more atoms of ar- 
 senic than a crystal of antimony contains atoms of antimony, 
 in the proportion of .0792 to .0519 ; if therefore we multiply, 
 for each element, the volume of one atom by the number of 
 atoms in a crystal of each, is it not perfectly clear that we ob- 
 tain the volume or bulk of all the matter, arsenic or anti- 
 mony, which each crystal contains ? On submitting our rea- 
 soning to the rigid test of calculation we actually find that a 
 crystal of antimony and an equal crystal of arsenic contain the 
 same volume of matter, for .0519 x 19.22 = .997518 and.0792 x 
 12.68 = 1.004256 the difference being only .006738, a coinci- 
 dence which must be regarded as very remarkable, when we 
 consider the unavoidable errors in the determinations both of 
 the specific gravities and the atomic weights. Still further, 
 we have said that tellurium is isomorphous with arsenic and 
 antimony ; now the atomic volume of tellurium is .978 and 
 
 * The sp. gr. of water is here taken as unity and not as before 1000. 
 
 3* 
 
^0 CttEJNflCAL CLASSIFICATION. 
 
 its atomic number 10.22, multiplying these two together we 
 obtain 9.99516 which is almost identical with the numbers 
 already obtained for antimony and arsenic ! Thus we have 
 
 Arsenic = 1.004256= unity very nearly. 
 
 Antimony =.9 9 75 18 = " 
 
 Tellurium=. 999516 = " 
 
 Mean =1.000430= « 
 
 We might perhaps apply to the numbers thus obtained, viz : 
 
 by multiplying the atomic volumes by the atomic numbers, 
 
 the term "crystalline volumes." Yet it will be immediately 
 
 obvious, from a consideration of the manner in which the 
 
 atomic numbers and the atomic volumes are obtained, that 
 
 their product must always be unity, since if W represent the 
 
 atomic weight of any substance and S its specific gravity, we 
 
 S 
 shall have for the atomic number — , and for the atomic vol- 
 
 W 
 
 w s w sw 
 
 ume — , now — x — = = !■ and since the same is true for 
 
 s' w s sw 
 
 all substances whatever, whether simple or compound, it fol- 
 lows that the crystalline volumes of all bodies are equal, that 
 is to say, every substance whatever, no matter what may be 
 its form or what its constitution, contains within a unit of vol- 
 ume, one cubic inch for instance, the same volume of ponder- 
 able matter. This may be stated otherwise as follows : " The 
 volumes of the atoms of all substances are inversely propor- 
 tional to the number of atoms which are contained under the 
 same surface." This law may be made of great use in the 
 determination of the atomic weights of different substances. 
 Its truth appears to us incontrovertible, if we admit the truth 
 of the premises. The members of group 2d are connected 
 with those of group 1 st, as already observed by the isomor- 
 phism of tellurium with arsenic and antimony. In addi- 
 tion to this, however, there are other points of resemblance 
 which are not unworthy of notice. Thus, the nitric, phos- 
 phoric, arsenic and antimonic acids are analogous in constitu- 
 tion to the iodic, chloric and bromic acids, while nitrous, ar- 
 
GROUP III. 31 
 
 senious, phosphorous and antimonious acids find parallels in 
 sulphuric, selenic and telluric acids. Hypo-nitric and hypo- 
 antimonic acids again are analogous in type to chlorous acid ; 
 N0 4 and Sb0 4 to C10 4 . Antimonious acid moreover resem- 
 bles the iodic in mode of combination as well as in constitu- 
 tion, thus we have KO 4- Sb0 5 and KO + 2Sb0 5 , corresponding 
 to KO + 10 5 and KO -f-2I0 5 . Arsenic and antimony, the former 
 especially, appear sometimes to replace sulphur in combination, 
 and to act the part of basacigen bodies. Thus we have FeS 2 + 
 FeAs— CoS 2 +CoAs — NiS 2 4-NiAs— Co 2 As 3 — FeAs— Ni 2 
 As — Mn 2 As — NiS 2 4-NiSb — Ni 2 Sb which we may assimi- 
 late to FeS 2 + FeS or Fe 2 S 3 — CoS 2 -f CoS or Co 2 S 3 — NiS 2 + 
 NiS or Ni 2 S 3 — Mn 2 S — Ni 2 S, where arsenic and antimony 
 appear to be substituted for an equivalent of sulphur in the 
 metallic sesquisulphides. 
 
 GROUP III. 
 
 The third group comprises also four elements. They are 
 Hydrogen, Zinc, Cadmium and Magnesium, which collec- 
 tively form an extremely well defined natural family. The 
 types of the compounds of these elements with the members 
 of the 1st or basacigen group are few and simple. If as usual 
 A represents a member of group 1st, and R any element be- 
 longing to the group now under consideration, we may com- 
 prehend all the combinations of the members of the two 
 groups under the general formulas, 
 
 R 2 A — RA — RA 2 , corresponding to which we have Zn o ? 
 — ZnO— Zn0 2 — HO— HS— H0 2 — HS 2 — CdO— MgO-ZnCl 
 — HC1— CdCl— MgCl, &c. &c. 
 
 The three peramphides or rather deut-amphides of this 
 family, viz : Zn0 2 — H0 2 — HS 2 are remarkable for the insta- 
 bility of their composition, and hence it is not very probable 
 that we shall ever obtain higher oxides, sulphides, &c. All 
 the proto-amphides of the members of this group are on the 
 other hand remarkable for stability and form very powerful 
 bases which enter into the compositfon of a great number and 
 variety of salts. The similar salts of the protoxides of zinc, 
 
32 CHEMICAL CLASSIFICATION* 
 
 cadmium and magnesium are isomorphous. Hydrogen is con- 
 nected with the others by numerous and striking analogies, 
 though it enters into the constitution of innumerable substan- 
 ces which find no parallels among the combinations of the 
 other members of the group. Recently, indeed, it has been 
 shown by Dr. Kane that hydrogen is to be regarded as a metal, 
 differing from other metals simply in physical constitution,being 
 at ordinary temperatures a gas, in the same manner and for the 
 same reason that mercury is at ordinary temperatures a liquid. 
 The analogies of hydrogen to zinc, are strikingly illustrated 
 by the following formulae first pointed out by Dr. Kane. 
 ZnCl + 6ZnO HC1+6HO 
 
 ZnCl+6ZnO + 6HO HC1 + 6HO + 6HO 
 
 ZnCl + 6ZnO + 10HO HC1 + 6HO + 10HO 
 Again, S0 3 + ZnO + 7HO S0 3 + ZnO + 2HO,5ZnO 
 
 S0 3 -fZnO+4HO,3ZnO S0 8 + ZnO + 7ZnO. 
 The equivalent weights of the members of this group are 
 not connected by any simple multiples, in other words they 
 are not isomeric. The atomic numbers of zinc and cadmium 
 are very different, that of the former being 21430, that of the 
 latter 15570. Their atomic volumes are consequently also 
 different. All the members of the group are in a high degree 
 electropositive; the replacement of an equivalent of hydro- 
 gen by an equivalent of zinc being the method universally 
 employed to generate voltaic currents. This group is not con- 
 nected with the second by any definite links, in the present 
 state of our knowledge, and its relations to group 1st are de- 
 pendant chiefly upon its intimate connexion with the next 
 family, with which indeed it is associated both by Graham 
 and by Kane, 
 
 GROUP IV. 
 
 Group 4th comprises Iron, Manganese, Chromium, Cobalt 
 and Nickel. The members of this group, like those of the 
 last, constitute an extremely well defined natural family. The 
 types of their combinations with the members of group 1st 
 are numerous, and it is in this respect that they differ from 
 
GROUP IV. 
 
 33 
 
 the elements constituting group 3d, with which, as already ob- 
 served, their connection is very intimate. These types are as 
 follows : 
 R 2 A— RA— RA 2 — RA 3 — R 2 A 3 --R 2 A 7 — R 2 A 5 — R 3 A 4 
 
 Corresponding to which we have 
 MnO. Mn 2 S-MnO-Mn0 2 -Mn0 3-Mn 2 03-Mn 2 7 -Mn 2 05-Mn 3 4 
 FeO. 1 FeO— FeS 2 — Fe0 3 — Fe 2 3 — 1 1 Fe 3 4 
 
 CrO. 1 CrO— 1 Cr0 3 — Cr 2 3 — ? 1 1 
 
 CoO. 1 CoO— C1S 2 1 Co 2 3 — 1 1 Co 3 4 
 
 NiO. Ni 2 S-NiO— NiS 2 ? Ni 2 3 — 1 1 1 
 
 In their general chemical relations the five members of this 
 group very strongly resemble each other. They all form 
 protoxides which are powerful bases and the similar salts of 
 which are isomorphous. The most interesting of these salts 
 is a class of sulphates constituted according to the general 
 formulae 
 
 RO + S0 3 + HO + 6aq 
 isomorphous, and capable of forming double salts with the 
 sulphates of soda and potassa, in which one equivalent of 
 either of these last replaces the one equivalent of constitu- 
 tional water in the original salt. The sesquioxides of the 
 members of this group form also by uniting with acids, sev- 
 eral very remarkable series of compounds, which like the salts 
 of the protoxides are isomorphous. Chromium, manganese 
 and iron form acids by uniting with three equivalents of oxy- 
 gen. Of these the chromic acid is the most stable, the man- 
 ganic acid less so, the ferric acid is decomposed by the slight- 
 est causes. The other two elements, cobalt and nickel, do not 
 form acids at all, so far at least as we at present know. Here 
 the law of grades is distinctly shown. Manganese again 
 forms two distinct acids with oxygen, to which the other mem- 
 bers of the group offer no parallel. The number of com- 
 pounds formed by manganese is moreover greater than the 
 number of those formed by any one of the other members of 
 the group. For these reasons we place manganese at the 
 head of the group, and may without impropriety give its name- 
 to the whole family. The members of this group unite with 
 cyanogen to form a remarkable series of compounds, which 
 
34 CHEMICAL CLASSIFICATION. 
 
 by their union as cyan-acids with cyano-bases, form a series 
 of salts, the elements of which are united by most powerful 
 affinities. Of these cyano-salts there are two classes ; in the 
 first class, a protocyanide of a metal of group four, acts as an 
 acid towards the cyanides of other radicals ; in the second 
 class of salts a sesquicyanide of a metal belonging to group 
 four, acts as an acid towards the cyanides of other radicals or 
 of other metals belonging to the same group. Belonging to the 
 first of these classes of salts we have but one series, viz : the 
 cyano-ferrites represented by the general formula 
 Fefcy+2MCy. 
 
 The second class of salts is more numerous and is repre- 
 sented by the formula 
 
 R 2 Cy 3 +3MCy. 
 
 Corresponding to which we have 
 Fe 2 Cy 3 +3KCy,Mn 2 Cy 3 +3KCy,Cr 2 Cy3+3KCy,Co 2 Cy 3 
 -f 3KCy, Ni 2 Cy 3 +3KCy, (?) 
 
 The members of this group are connected with those of group 
 1st, by several distinct and well ascertained isomorphisms. Thus 
 the chromic and manganic acids form salts which are isomor- 
 phous with the corresponding sulphates and seleniates. The 
 chromates moreover resembles the sulphates in solubility and 
 in their mode of forming double salts. Again, a peculiar acid 
 of manganese has occasionally been obtained in combination 
 with potassa, forming a salt, the formula of which is Mn 2 s 
 4-KO, and which is isomorphous wtth chlorate of potassa 
 C1CL+KO. Still farther, the salts of hypermanganic acid 
 are isomorphous with the corresponding salts of hyperchloric 
 acid Mn 2 7 + R0 with C10 7 + R0. Hence it is clear that, in 
 combination at least, two equivalents of manganese replace and 
 are isomorphous with one equivalent of chlorine. And thus 
 group 1st and 4th are connected by a two-fold isomorphism. 
 We may here observe that as one equivalent of manganese 
 is in combination isomorphous with one equivalent of sulphur, 
 two equivalents of manganese ought to be isomorphous with 
 two equivalents of sulphur, consequently the salts of manga- 
 nese of the type Mn 2 5 + RO, which from analogy we will 
 
GROtJ? V, 35 
 
 call hypomanganates, should be found isomorphous with the 
 hyposulphates, S 2 5 + RO, and the hyposulphates should 
 therefore be isomorphous with the chlorates. We have no- 
 ticed this point because the attention of chemists does not 
 seem to have been directed towards it, obvious as it may ap- 
 pear. The salts of the protoxides of this group are isomor- 
 phous with the similar salts of the protoxides of group 3d, 
 and there is furthermore a very great analogy in other respects 
 between the salts of the two series of oxides. All the metals 
 of group 4th dissolve in hydrated acids by a simple replace- 
 ment of hydrogen, and the three types of the compounds of 
 group 3d correspond to the first three types of the compounds 
 of group 4th, with the basacigen family. 
 
 The metals composing this group possess nearly the same 
 atomic weights and specific gravities, and further researches 
 may perhaps identify them in these respects. The specific 
 gravity of chromium however differs greatly from the mean 
 specific gravity of the other four metals. The mean of the 
 atomic members of iron, cobalt, manganese and nickel, is 
 
 .2896, and the mean atomic volume is consequently ■ or 
 
 .2896 
 
 3.4530. It is worthy of notice that the molecules pi these 
 four metals are smaller than those of any other simple sub- 
 stance with which we are acquainted. 
 
 Finally, iron and nickel are the only two substances which 
 are certainly known to be susceptible of magnetism, though 
 cobalt and manganese have been supposed to share this pro- 
 perty also ; the equivalents of the two former we may also 
 remark have the same specific heat. 
 
 GROUP V. 
 
 The fifth group comprises Aluminum, Glucinum and Zirco- 
 nium. These three elements are nearly related to those of 
 group four, and are classed with them by Graham and by- 
 Kane. They differ however essentially from them in forming 
 but a single molecular type by combining with the members 
 of group 1st. This type is represented by the formulge R 2 A 3 . 
 
35 CHEMICAL CLASSIFICATION, 
 
 The sesquioxides, alumina, glucina and zirconia, are capable 
 of acting the part of acids toward metallic protoxides as well 
 as of powerful bases toward the strong acids. The salts of alu- 
 mina and glucina are probably isomorphous with those of 
 zirconia, though this point has not yet been satisfactorily as- 
 certained, they are however certainly isomorphous with the 
 sesquioxides of the members of group four, both in a free state 
 and in a state of combination, Alumina frequently replaces 
 the sesquioxides of iron in mineral compounds as well as in 
 artificially prepared salts. ^-As an instance of the striking 
 analogy between the chemical relations of the sesquioxides 
 of groups four and five we may cite the general formulae of 
 the alum family. 
 
 R'0,S0 3 +R 2 3 ,3S0 3 -{-2dag, 
 corresponding to which we have the alumina, iron, manga- 
 nese and chrome-alums. In like manner we have a class of 
 double oxalates, the general formulae of which is 
 3(R'0,C 2 3 ) + R 2 3 .3C 2 3 +6ag, 
 corresponding to which we have double oxalates of chro- 
 mium, aluminum, and iron with oxalates of potassa, which are 
 like the alums isomorphous. The elements which we have 
 arranged under group 5th are connected with the basacigen 
 family, and with the members of group 3d, indirectly by their 
 relation with group 4th. The three last groups, viz : groups 
 3d, 4th, and 5th, constitute in a great measure the 2d or mag- 
 nesian class of Graham. They are, as we have already stated, 
 very intimately connected, as well by chemical as by physi- 
 cal properties, though we have thought it more in accordance 
 with the true principles of a natural classification, to arrange 
 them under three separate classes or groups. As an illustra- 
 tion of their capacity of replacing one another in combination, 
 we cite the following formulae derived for the most part from 
 the mineral kingdom. 
 
 Fe0 + Fe 2 3 . Zu0 + Al 2 3 . MnO-fMn 2 3 . 
 
 FeO + Cr 2 3 . MqO + Al 2 3 . Zuo + Mn 2 O s . 
 We may remark in conclusion that aluminum is isomeric 
 with iron and manganese 13.72 x 2=27.44, and that the equiv- 
 
alent Of glucinum so nearly approaches that of the other mem- 
 bers of the two groups, viz : the 4th and 5th, that we may 
 reasonably expect that upon more accurate investigation, it 
 will be found isomeric with them. The equivalent of zirco- 
 nium is however entirely removed from the series and approx- 
 imates toward that of zinc. 
 
 GROUP VI. 
 The sixth group comprises four elements, Molybdenum, 
 Tungsten, Vanadium, Columbium. The compounds of these 
 elements with the members of group 1st may be arranged un- 
 der three types. • 
 
 RA— RA 2 — RA 3 
 corresponding to which we have 
 
 MoO--Mo0 2 — Mo0 3 
 ? TnQ 2 — Tn0 3 
 VO— VO,-V0 3 
 ? W0 2 -W0 s 
 It will be immediately observed that these types are, so far 
 as they extend, precisely similar to those of group 4th. Sev- 
 eral' of the salts of tungstic and molybdic acids are isomor- 
 phous, while on the other hand, one of the forms of chromate 
 of lead corresponds with that of the molybdate of the same 
 base. Between the vanadic and chromic acids there is also 
 considerable resemblance, though their compounds are not 
 known to be isomorphous. There exist, moreover, compounds 
 of Molybdenum and Tungsten with oxygen and chlorine, pre- 
 cisely analogous in composition to the so-called chloro-sulphu- 
 ric and chloro-chromic acids, thus we have 
 2Cr0 3 +CrCl 3 or Cr0 2 ,CL 
 2S0 3 +SC1 3 or S0 2 ,C1. 
 2Mo0 3 +MoCl 8 *or Mo0 2 ,Cl. 
 2WO 4-WCI3 or W0 2 CI. 
 Formulae, which together with the analogies which we have 
 already mentioned, connect this group very evidently with 
 the 4th, and indirectly with the 1st, 3d, and 5th groups. The 
 protoxides of Vanadium and Molybdenum are not powerful 
 
 4 
 
38 CHEMICAL CLASSIFICATION. 
 
 bases like those of group 4th, neither do the four metals com- 
 posing this group form with the basacigen family sesqui-com- 
 pounds as do all the members of group four. Molybdenum 
 forms with sulphur a higher acid than the sulpho-molybdic, 
 the formulae being MoS 4 , and thus apparently connects this 
 group with group 2d ; the compounds of this acid with sul- 
 phur-bases are well worthy of attention, only one compound,, 
 the potassium-salt MoS 4 + KS, being at present known. The 
 chemical properties and^rejations of the four elements which 
 we have associated together as group 6th are in general very 
 similar, though further researches are wanting to enable us to 
 assign with certainty the proper position of columbium. 
 
 Both the atomic weight and the specific gravity of Tung- 
 sten are very nearly double those of molybdenum, thus, 
 2Mo=95.92=W=94.8 nearly, and S.GW=17.22=2S.GMo 
 = 17.23. The atomic numbers or chemical densities are con- 
 sequently the same for both metals, and they likewise agree 
 in atomic volume, or in the size of their molecules. The mean 
 atomic number is .1811, and the mean atomic volume 5.52. 
 The specific gravities of vanadium and columbium are as 
 yet unknown. 
 
 GROUP VII. 
 
 Group 7th comprises Copper, Mercury, Bismuth, and Pal- 
 ladium. The types of the compounds formed by the mem- 
 bers of this group with those of group 1st are three in num- 
 ber. 
 
 R 2 A— RA— RA 2 , 
 corresponding to which we have 
 
 Cu o 0— CuO— Cn0 2 
 
 Hg~ 2 0— HgO— ? 
 
 Bi 2 0?— BiO— Bi0 2 
 
 Pd 2 0— Pd0—Pd0 2 
 In their general relations the four metals composing group 
 7th, agree in a remarkable manner, though there exist between 
 their compounds no observed isomorphisms. Their corres- 
 ponding oxides are all insoluble in water ; the suboxides of 
 
group vii. 39 
 
 copper, mercury and palladium forming definite and exceed- 
 ingly well marked salts with acids, a property, it may be re- 
 marked, which is rare in oxides of that constitution. The 
 protoxides of each is in like manner a powerful base, while 
 the pertoxides of copper, bismuth and palladium are analo- 
 gous in instability and in incapacity to form well denned com- 
 pounds with either acids or bases. The protoxides of cop- 
 per, bismuth and mercury exhibit moreover a remarkable 
 tendency to form compounds containing an excess of base. 
 The subnitrates of copper and bismuth are analogous in 
 composition, according to Graham, being — N0 5 ,HO-i-3CuO 
 and NO § ,HO + 3BiO ; their subsulphates also agree in compo- 
 sition if we omit basic water, thus, S0 3 +3CuO and S0 3 + 
 3BiO. Group 7th is connected with groups 3d and 4th by 
 the isomorphism of the double sulphate of copper and po- 
 tassa with the corresponding double sulphates of zinc and 
 potassa, and of iron and potassa, thus we have CuO,S0 3 + 
 KO,S0 3 +6aq isomorphous with ZnO,S0 3 +KO,S0 3 -f 6aq 
 and FeO,S0 3 +KO,S0 3 +6aq. The double cyanide of mer- 
 cury and potassium is moreover isomorphous with the double 
 cyanide of zinc and potassium HgCy-fKCy with ZnCy + 
 KCy, both these salts appearing in octohedrons of the regular 
 system. Bismuth is associated with the 2d or nitrogen group 
 by Dr. Kane, on the ground of the isomorphism of its sulphide 
 with the tersulphide of Antimony, yet as its oxides do not 
 form well defined salts with alkaline bases, and as it does not 
 form with hydrogen the characteristic gaseous compound of 
 the type RH 3 , we cannot admit the propriety of displacing it 
 from its original position. 
 
 No isomerism exists between the equivalents of the four el- 
 ements composing this group, nor are their atomic numbers 
 and atomic volumes connected by any very certain and defi- 
 nite relations. The atomic numbers of the four metals are 
 as follows: Copper=.2742, Mercury .1321, Bismuth .0922, 
 Palladium .2153, taking the number for Copper as unity, the 
 number for the other metals will be to it nearly in the follow- 
 
49 CHEMICAL CLASSIFICATION 
 
 ing ratios, mercury |, bismuth J, palladium J, and the atomic 
 volumes will of course be in the inverse ratio, viz : copper 1, 
 mercury 2, bismuth 3, palladium If. 
 
 GROUP VIIL 
 
 The 8th group of our classification comprises 5 elements, 
 viz: Barium, Strontium, Calcium, Lead and Silver. The 
 types of the compounds of the members of this group with 
 the members of group ls^are three in number. 
 
 R 5 A— RA— RA 2 , 
 corresponding to which we have 
 
 Ag,0— AgO— Ag0 2 
 Pb 2 0— PbO— Pb0 2 . 
 
 ? _BaO— Ba0 ; 
 ? _SrO— Sr0 2 .' 
 ? _CaO— CaO 
 
 2' 
 
 The similarly constituted salts of barium, strontium and lead 
 are isomorphous and agree remarkably in solubility in water 
 and in their purely chemical relations. Thus the sulphates of 
 these three elements occur in nature crystallized in the same 
 form, and when artificially prepared agree in their insolubility 
 in appearance, and in their chemical properties. In like man- 
 ner the nitrates of lead, barium and strontium are isomor- 
 phous and similarly constituted. Calcium is connected with 
 the other members of this group, by the isomorphism of one of 
 the forms of its carbonate, viz : arragonite, with the carbonates 
 of baryta and strontia, by the insolubility of its sulphate, and 
 by a general similarity in chemical relations. We may also 
 add, that agreeably to the researches of Newmann, the speci- 
 fic heats of the equivalents of the Carbonates of Lime, Baryta? 
 strontia and lead are the same, and that the same relation 
 of equality holds good for the specific heats of the sulphates 
 of the same basis. Silver was first classed with lead by Dr° 
 Kane, and the two metals agree very closely in chemical pro- 
 perties, in the insolubility of their chlorides, bromides and 
 iodides, &c, but their salts are not connected by any known 
 isomorphisms. On the whole then we may regard this group 
 
GROUP IX. 41 
 
 as being extremely well defined and as presenting a fair ex- 
 ample of a Natural family in Chemistry. With groups 3d and 
 4th this family is connected by the isomorphisms of the carbon- 
 ates of strontia and baryta, with one of the forms of the carbon- 
 ates of magnesia and of iron, by the isomorphism of the ordinary 
 carbonate of lime or Iceland spar with the usual forms of the 
 carbonates of iron and magnesia, and by the analogy in con- 
 stitution and in form of sulphates of iron and of lime whose 
 formulae are S0 3 + FeO + 2HQ andCaO + S0 3 +2HO. The 
 relations between this group and group 1st are in the present 
 state of our knowledge indirect, being traced chiefly through 
 groups 3d and 4th. 
 
 No isomerisms subsist between the equivalents of the ele- 
 ments constituting this group. The atomic numbers and 
 atomic volumes of silver and of lead, the only members of the 
 group for which they can at present be determined, are also 
 unconnected by any simple ratios. The specific heat of the 
 atom of silver is double that of lead, and the former metal is 
 apparently more highly electro-negative than the latter. 
 
 GROUP IX. 
 
 Group 9th comprises Platinum, Titanium, Iridium and 
 Osmium. The complete series of the compounds formed by 
 the members of this group with those of group 1st are 5 in 
 number, viz : 
 
 RA— RA 2 — RA 3 — RA 4 — R 2 A 3 , 
 corresponding to which we have 
 
 PtCl— PtCl 2 — ? — ? — ? 
 
 TiO— TiCl a — ? — ? — ? 
 
 IrO— IrCl 2 — Ir0 3 — ? — Ir Cl, 
 
 OsCl— 0sCl 2 — OsCl 3 — 0s0 4 — 0s 2 CL. 
 The four elements which we have arranged under group 
 9th, correspond for the most part very closely in chemical pro- 
 perties and relations. The bichlorides of all these metals form 
 with chloride of potassium, highly characteristic chloro-salts 
 which crystallize in regular octahedrons. Their general 
 
42 CHEMICAL CLASSIFICATION. 
 
 formula is RCU+KC1. The corresponding compounds of 
 platinum, iridium and osmium are all isomorphous. None 
 of the protoxides of this group are very powerful bases and 
 their binoxides like compounds of that type generally, are ca- 
 pable of uniting to a certain extent with both acids and bases. 
 
 With the 7th, or copper family, this group is connected by 
 the isomorphism of the corresponding salts of the chloropallad- 
 ious, and chloropalladic^acids with the chloroplatinous and 
 chloroplatinic acids. 
 
 With group 2d this group is connected by the analogy in 
 composition of osmic acid with hypo-nitric and hypo-anti- 
 monic acids. With group 4th this group may possibly be 
 connected by the analogy of the sesquichlorides of iridium 
 and osmium to the sesquichlorides of iron and manganese. 
 
 Of the four metals composing this group platinum and iri- 
 dium are certainly isomeric, the equivalent of each being 98.84. 
 The equivalent of osmium, viz : 99.72, nearly approaches that 
 of platinum and iridium as to lead to the suspicion that more 
 accurate researches will establish their identity. Finally the 
 equivalent of Titanium as at present received is almost exact- 
 ly one-fourth of the equivalent of Platinum and Iridium — 24.5 
 X4=98.00. 
 
 The atomic numbers of platinum, iridium, and titanium 
 are also equal, the mean being .2179; the atomic volumes 
 consequently also correspond, being for each 4.59. The spe- 
 cific gravity of osmium is stated by Graham as being " about 
 10," in this case its atomic volume will be nearly twice and 
 its atomic number nearly one-half the atomic volumes and 
 atomic numbers of the other metals composing group 9th. 
 
 GROUP X. 
 
 Group 10th comprises Gold, Uranium, Rhodium and Tin. 
 The types of the compounds formed by the members of this 
 group with the first or basacigen family are four in number, 
 viz : R 2 A — RA— RA 2 — R 2 A 3 corresponding to which we have 
 Au 2 0, Au 2 3 , UO— U,6 3 — R 2 3 — SnO— Sn0 2 — Sn 3 0,. ' 
 
group x. 43 
 
 The sesquioxides of all these metals agree in being feebly acid 
 and basic; that of tin has as yet been but imperfectly examined. 
 The sesquichlorides of gold, uranium and rhodium, form 
 well characterized chloro-salts with the chlorides of the alka- 
 line metals. This group is connected with the last by the 
 isomorphism of the chlorostannates and chloroplatinates of 
 potassium, SnCl 2 + KC1 and PtCl 2 +KC1 ; by the isomorphism 
 of the native dentoxides of tin and titanium, Sn0 2 and Ti0 2 , 
 and by the analogy of the sesquichlorides of gold, uranium 
 and rhodium, with those of osmium and iridium. 
 
 Between the elements composing this group no probable 
 isomerism exists, though the atomic weight of gold, as stated 
 by Graham, is nearly identical with that of osmium, as at pre* 
 sent received. The atomic number of gold is identical with 
 that of silver, with which metal gold is isomorphous in the 
 metallic state ; but the analogy between the two metals ceases 
 here. 
 
 The atomic number of rhodium closely approximates to 
 that which we have stated for the last group, with which 
 therefore it is in this way connected. The atomic number 
 of uranium is remarkable as being very small, viz : .0414 ; its 
 atomic volume is consequently larger than that of any metal 
 yet described, viz : 24.15. 
 
 Finally the specific heats of the equivalents of gold and 
 tin agree with each other and with that of platinum. 
 
 GROUP XI. 
 
 The eleventh group of our classification comprises Potas- 
 sium, Sodium, and Lithium, commonly known as the alkaline 
 metals. These 3 elements are connected together by very 
 strong analogies in their chemical properties and relations, but 
 not by any well established isomorphisms. Their isomor- 
 phism in the metallic state appears tolerably certain. The most 
 energetic affinities are exerted by the protoxides of these 
 metals ; the greater number of the salts which they form with 
 oxy-acids are soluble in water and neutral in composition, 
 
44 CHEMICAL CLASSIFICATION. 
 
 and their relations to other oxides are almost invariably those 
 of bases to acids. With group 8th they are connected by the 
 isomorphism of the sulphate of soda with sulphate of silver. 
 Under certain circumstances an equivalent of potash or soda 
 may be replaced in combination by an equivalent of Lime-f 
 an equivalent of waj^r, or CaO + HO, and thus this group is 
 connected with group 8th and 4th at the same time. Again 
 we find that one equivalent of nitrogen + four equivalents of 
 hydrogen are capable of replacing potassium in combination 
 without change of form, NH 4 being isomorphous with K, and 
 thus this group is connected with Groups 2d and 3d. 
 
 No isomerism appears to subsist between the equivalents of 
 the members of this group. The atomic number of potassium 
 is very nearly double that of sodium, and its atomic volume is 
 consequently one half that of the latter metal. The specific 
 gravity of lithium appears not to have been determined, its 
 atomic weight is very remarkable as being lower than that 
 of any other metal, unless indeed hydrogen be regarded as 
 such. 
 
 GROUP XII. 
 
 The twelfth group comprises 2 elements— Yttrium and 
 Thorium. The compounds of these two metals are as yet 
 very imperfectly known. Their oxides are regarded as pro- 
 toxides, and bear considerable analogy to each other in chemi- 
 cal properties and relations. No definite relations have been 
 established between them and the members of any other group. 
 
 GROUP XIII. 
 Group 13th comprises Cerium, Lantanium, Didymium, Er- 
 bium and Terbium. Our knowledge of these substances is as 
 yet too much imperfect to admit of our entering into their rela- 
 tions and properties, sufficiently to establish their position in 
 the same class. They all occur in nature in the same minerals 
 and are exceedingly difficult to separate from one another, 
 by chemical means. The last four in the order which we 
 
GROUP XIV. 45 
 
 have mentioned, have been discovered by Mosander within a 
 few years, and are yet under his investigation. 
 
 GROUP XIV. 
 
 Group 14th comprises Carbon, Boron and Silicon. These 
 elements are associated together by Graham upon the ground 
 that they exhibit a general resemblance to one another with- 
 out any precise relations. Independently, however, of the 
 non-agreement of the types of their compounds with the basa- 
 cigen group, carbon enters into the constitution of an immense 
 number and variety of compounds, to which boron and silicon 
 offer no parallels. If however we reduce the received equiv- 
 alents of boron and silicon by one-third, as proposed in the 
 case of silicon by Dr. Clarke, the silicic and boric acids be- 
 come analogous in constitution to carbonic acid, S0 o and BO., 
 to C0 2 . This view of the constitution of the two former acids 
 is supported by their general analogy to carbonic acid in 
 chemical relations, they are feeble acids, and are at the same 
 time capable of entering into combination with bases in many 
 proportions. Moreover, the formulae of the compounds of 
 boron and silicon generally, become much more simple by 
 making the proposed change, than they are at present, as will 
 readily be seen from a comparison of the following formulae, 
 which we have calculated for the purpose. 
 
 Old View. 
 
 . 2SiF 3 +3KFl, 
 
 2BF 3 +3KF1, 
 
 3NaO + 2B0 3 , 
 
 Biborate of Soda Borax. NaO -f 2BO 3 , 
 
 NaO+ B0 3 , 
 Boracite. 3MgO + 4BO g , 
 
 Picrosmine. 3MgO + 2SiO 3 , 
 
 Pyroxene. 3MgO,2SiO 3 + 3CaO,2SiO , 
 
 Steatite. 3MgO,2Si0 3 + Al 2 3 2Si0 8 6HO 
 
 Amphigen or Analcime. 3RO,2SI0 3 + 3 (Al 2 35 2Sio 3 )±aq 
 Felspar, Albite, Stilbite. RO ; Si0 3 + A) a 3 3Si0 3 ±aq. 
 
46 chemical classification. 
 
 New View. 
 SiF 2 + KF, (Clark.) 
 BF 2 + KF, 
 Na0+B0 2 
 Biborate of Soda Borax. NaO + 3B0 2 
 
 ^ 2NaO+3B0 2 like2NaO + 3C0 2 . 
 Boracite. MgO-f2B0 2 
 
 Picrosmine. M gO + SiO 2 
 
 Pyroxene. MgO,Si0 2 + CaO,Si0 2 like MgO, 
 
 C0 2 +CaO,C0 2 . Dolomite. 
 Steatite. Mg0,Si0 2 +Al 2 3 ,3Si0 2 +6H0 
 
 Amphigen or Analcime. R0,Si0 2 + 2 (Al 2 3 ,3Si0 2 )±aq 
 Felspar, Albite, Stillite. R0,3Si0 2 + Al 2 ~0 3 3Si0 2 ~ 
 
 The formulae of other borates and silicates are not however 
 always rendered more simple by the proposed change in the 
 equivalents of boron and silicon, so that the expediency of 
 such a change must remain for the present undecided, and we 
 must admit, that, as in the case of the last group, further re- 
 searches are wanting, before we can assign with any degree 
 of probability, the proper places of the elements, carbon, boron 
 and silicon. 
 
 Thus far then we have endeavoured to display the analo- 
 gies which subsist between the elementary bodies. We have 
 divided them into natural families or groups, according to our 
 idea of the true principles which must govern every attempt 
 to create a natural classification in any branch of science what- 
 ever. In a word, we have endeavoured to rely upon general 
 and not upon particular, analogies, and our classification con- 
 sequently differs for better or for worse from every one which 
 has yet been proposed. The merit of having first suggested 
 such a mode of classification belongs to the English chemist, 
 Graham, who, in his recent treatise upon Chemistry, has divi- 
 ded the elements into eleven classes, depending chiefly upon 
 their isomorphous relations. We will here quote the classes 
 of Graham in a column over our own. 
 
GROtrp xiv. 4 7 
 
 Graham. 
 1st Class. Oxygen, Sulphur, Selenium, Tellurium 
 2d Class. Mg— Ca— Mn— Fe— Co— Ni— Zn— Cd— Cu— H— - 
 
 Bi— Cr— Al— Gl— V— -Zn— Y— Th. 
 3d Class. Barium, Strontium, Lead. 
 4th Class. Potassium, Sodium, Ammonium, Silver. 
 5th Class. Chlorine, Iodine, Bromine, Fluorine. 
 6th Class. Nitrogen, Phosphorus, Arsenic, Antimony. 
 7th Class. Tin, Titanium. 
 8th Class. Gold, Silver. 
 
 9th Class. Platinum, Palladium, Iridium, Osmium. 
 10th Class, Tungsten, Molybdenum. 
 11th Class. Carbon, Boron, Silicon. 
 Unclassified. Mercury, Cerium, Columbium, Rhodium, Uran- 
 
 ium, Lanthanum, (Erbium and Terbium.) 
 
 1st Class. Oxygen, Sulphur, Selenium, Tellurium, Chlorine. 
 Iodine, Bromine, Fluorine. 
 
 2d Class. Phosphorus, Nitrogen, Arsenic, Antimony. 
 
 3d Class. Hydrogen, Zinc, Cadmium, Magnesium. 
 
 4th Class. Iron, Manganese, Cobalt, Nickel, Chromium. 
 
 5th Class. Aluminum, Glucinum, Zirconium. 
 
 6th Class. Molybdenum, Tungsten, Vanadium, Columbium, 
 
 7th Class. Copper, Mercury, Bismuth, Palladium. 
 
 8th Class. Barium, Strontium, Calcium, Lead, Silver. 
 
 9th Class. Platinum, Titanium, Iridium, Osmium. 
 
 10th Class. Gold, Uranium, Rhodium, Tin. 
 
 11th Class. Potassium, Sodium, Lithium. 
 
 12th Class. Yttrium, Thorium. 
 
 13th Class. Cerium, Lanthanum, Didymium, Erbium, Ter- 
 bium. 
 
 14th Class. Carbon, Boron, Silicon. 
 We are likewise indebted to the suggestions of Dr. Kane, 
 
 for several important changes in the original classification of 
 
 Graham : thus lead and silver Were first associated together 
 
 by the Irish chemist, as were also palladium and copper. 
 
4S CHEMICAL CLASSIFICATION. 
 
 With certain other changes proposed by Kane we do not co- 
 incide, as they are entirely opposed to our convictions of 
 truth. 
 
 The tendency of discovery, in our time, is to connect the 
 different groups arflfthe individual members of which they 
 are composed more and more closely together. Every year 
 fills up some separating chasm, brings to light new analogies, 
 and impresses us yet more strongly with the belief that nature 
 in truth, is one in essence, though many in form. Perhaps it 
 will hereafter be found that those bodies which we term ele- 
 ments are in reality compound ; that they are the resultants of 
 a few elementary forces, themselves perhaps but modifications 
 of a single primary force. The very idea of matter, reduced 
 to its simplest form, is the idea of resistance, and the idea of 
 resistance, think of it as we may, is only one form of the idea 
 of force. To us and to our nature, an atom is a sphere of force 
 to whose centre we cannot get. This view of the constitu- 
 tion of matter, will at least afford us some explanation of those 
 analogies which now seem so difficult to comprehend ; for the 
 resultants of the same forces must of necessity bear some re- 
 lation and proportion to one another ; and it seems to us at 
 least, that the assumption of a few primary forces, whose in- 
 numerable modifications and modes of action, produce all the 
 phenomena of material bodies with which we are acquainted, 
 enables us to form a simpler and a clearer idea of the unity of 
 nature than any hypothesis which has yet been proposed. 
 
 To exemplify at a glance the relations of the several groups 
 to each other, and to show, at the same time, in what manner 
 the elements form a natural scale of gradation, we propose to 
 make use of an artifice often adopted by naturalists, and for 
 the suggestion of which the writer is indebted to Dr. Torrey. 
 
 Assuming a single element, oxygen for example, as a cen- 
 tre, we may draw around it a circle of any convenient radius, 
 and we may divide this circle into segments, each segment 
 corresponding to some particular point of analogy. 
 
 Concentric with this primary circle, and separated from it by 
 
SCHEME OF AXilOfjlES. 
 
 ITH. OF S* W. END1COTT. 
 
COMPOUNDS. 49 
 
 a short interval of radius, we may draw another which shall 
 serve to include those substances which are connected with 
 oxygen and with each other by the first degree of affinity, 
 and we may place each substance opposite the segment of the 
 primary circle which indicates the point of analogy by which 
 it is most closely connected with oxygen. A third circle, ex- 
 ternal to, and concentric with the last, will include those 
 bodies which are connected with oxygen by the second de- 
 gree of analogy, and so on for other circles. Blank segments, 
 or blank circles will indicate want of analogy in particular 
 points or in entire degrees. And thus we shall obtain a 
 graphical illustration of the connection of the elements with 
 each other, and of the degree of analogy between the several 
 groups. 
 
 OF THE CLASSIFICATION OF COMPOUNDS. 
 
 The classification of compounds, according to the natural 
 system, obviously depends upon the previous classification of 
 the elements, and follows readily and easily from a knowledge 
 of the arrangement and relations of the elementary groups. 
 Upon this portion of our subject we propose to be as brief as 
 is consistent with clearness. The principles of classification 
 which we shall adopt are as follows : 
 
 1. The compounds of the members of group 1st with one 
 another are arranged according to their molecular types. 
 
 2. The compounds of the members of any other group than 
 group 1st with the members of group 1st, are arranged to- 
 gether in groups according to molecular types, principally and 
 according to isomorphism, wherever isomorphisms subsist. 
 
 For example, we classify together the Protoxides, Proto- 
 sulphides, Proto-selenides, Proto-tellurides, Proto-chlorides, 
 Protiodides, Proto-bromides, and Proto-fluorides of the mem- 
 bers of each natural group of elements. Supposing, for ex- 
 ample, that a particular natural group contains four elements, 
 we should have of course a group of proto-amphides of this 
 elementary group, consisting of 4 x 8 = 32 individuals. These 
 
 5 
 
50 CHEMICAL CLASSIFICATION. 
 
 again we subdivide into eight sub-groups, each sub-group con- 
 sisting (in this particular instance) of four individuals. The 
 sub-groups are arranged according to their electro-negative or 
 chlorous constituent, so that we thus associate together in the 
 same sub-group, all the protoxides of a particular class of el- 
 ements ; in another sub-group all the proto-sulphides of the 
 same class of elements, and so on. 
 
 The same principle is applied to the classification of deuto- 
 amphides, sesqui-amphides, and so forth. 
 
 As examples of this method of arrangement, we will give 
 a few groups and sub-groups of different molecular types. 
 
 THIRD GROUP OF ELEMENTS. 
 
 Zinc, Cadmium, Magnesium, Hyd7 , ogen. 
 
 GROUP OF PROTO-AMPHLDES. 
 
 Sub-group 1st. Sub-group 2d. Sub-group 3d. Sub-group 4th. 
 ZnO ZnS ZnSe ZnTe 
 
 CdO CdS CdSe CdTe 
 
 MgO MgS MgSe MgTe 
 
 HO HS HSe HTe 
 
 Sub-group 5th. Sub-group 6th. Sub-group 7th. Sub-group Sth. 
 ZnCl Znl ZnBr ZnF 
 
 CdCl Cdl CdBr CdF 
 
 MgCl Mgl MgBr MgF 
 
 HC1 HI HBr HF 
 
 As the members of the 3d group of elements do not form 
 very distinct compounds with more than one equivalent of a 
 member of the first or amphigen group, we must refer to some 
 other one of the elementary groups for further illustrations. It 
 will be immediately observed in the example which we have 
 already given, that the Protoxides of group 1st constitute to- 
 gether a natural family or sub-group, the protosulphides a sec- 
 ond sub-group and so on : hence in a descriptive work, the 
 sub-groups should be treated of in succession, first by a gen- 
 eral account of their properties and relations, and secondly by 
 
COMPOUNDS. 51 
 
 the special description of the individuals of which each sub- 
 group is composed. 
 
 Another example of a natural sub-group is furnished by the 
 sesquioxides of the fourth group of elements, viz : 
 Fe 2 3 ,Mn 2 3 ,Cr 2 3 , Co 2 3 , Ni 2 3 . 
 
 Again we have as a sub-group of another type, 
 N0 5 , P0 5 , As0 5 , Sb0 6 . 
 
 Again, PbCl 2 , IrCl 2 , 0sCl 2 , TiCl 2 . 
 
 From this principle of classification it follows that we shall 
 have for each group of elements, several groups of com- 
 pounds, with the members of group 1st, the number depend- 
 ing on the number of the different types of the compounds 
 formed by the two groups with each other. 
 
 Thus, if a particular group of elements form with the 1st 
 or Amphigen group, compounds of the types, RA, RA 2 , 
 R 2 A 3 , RA 3 , it is clear that we shall have four groups of com- 
 pounds, namely : 
 
 1. Proto-amphides RA 
 
 2. Deuto-amphides RA 2 
 
 3. Sesqui-amphides R 2 A, 
 
 4. Ter-amphides RA 3 
 
 Each of these groups will then be subdivided into eight 
 sub-groups, as follows : 
 
 Group 1st, Proto-amphides. Group 2d, Deuto-amphides. 
 
 Sub-group 1st, Protoxides RO 1st, Deuotoxides R0 2 
 
 Sub-group 2d, Proto-sulphides RS 2d, Deuto-sulphides RS 2 
 
 Sub-group 3d, Proto-selenides B,Se 3d, Deuto-selenides RSe 2 
 
 Sub-group 4th, Proto-tellurides RTe 4th, Deuto-tellurides RTe 2 
 
 Sub-group 5th, Proto-chlorides RC1 5th, Deuto-chlorides RC1 2 
 
 Sub-group 6th, Prot-iodides RI 6th, Deut-iodides RI 2 
 
 Sub-group 7th, Proto-bromides RBr 7th, Deuto-bromides RBr 2 
 
 Sub-group 8th, Proto-fluorides RF 8th, Deuto-fluorides RF 2 
 
 Group 3d, Sesqui-amphides. Group 4th, Teramphides. 
 
 Sub-group 1st, Sesqui-oxides Ro0 3 1st, Teroxides R0 3 
 
 Sub-group 2d, Sesqui-sulphides R 2 S 3 2d, Tersulphides RS 3 
 
 Sub-group 3d, Sesqui-selenides R 2 Se3 3d, Terselenides RSe3 
 
 Sub-group 4th, Sesqui-tellurides R 2 Te 3 4th, Tertellurides RTe 3 
 
52 CHEMICAL CLASSIFICATION. 
 
 Sub-group 5th, Sesqui-chlorides R 2 CI 3 5th, Terchlorides RCI - 
 Sub-group 6th, Sesqui-iodides RoI 3 6th, Teriodides Rig 
 Sub-group 7th, Sesqui-bromides R 2 Br 3 7th, Terbromides RBr g 
 Sub-group 8th, Sesqui-fluorides R 2 F 3 8th, Terfluorides RF 3 
 
 3. Compounds which do not contain an element of the am- 
 phigen group are classified together, according to their consti- 
 tution and general relations, into groups. Examples, 
 
 NH 3 — PH 3 — AsH 3 — SbH s 
 P 2 Cu 3 — Po°Co 3 — P 2 Ni 3 — P 2 Mn 3 . 
 With the exception of ammonia, the carburets of hydro- 
 gen, and one or two other substances which we shall here- 
 after notice, our knowledge of the compounds of the non- 
 amphigen groups is as yet too imperfect to admit of our en- 
 tering upon any generalizations concerning them. 
 
 4. Before entering upon the subject of the classification of 
 double compounds, it is necessary to make a few remarks, 
 concerning the substitutions of the elements generally, and 
 more particularly concerning the substitutions of the members 
 of the amphigen group for one another. 
 
 We admit then generally that the members of any group 
 are capable of replacing one another in equivalent propor- 
 tions, without destroying the type of the particular compound 
 in which the replacement occurs, though the chemical relations 
 of the new compound may or may not be the same with those 
 of the original from which it is derived. We therefore clas- 
 sify compounds thus derived by substitution with those from 
 which they are derived, provided that the chemical relations 
 of the primitive and its derivatives are not absolutely incom- 
 patible. Examples. 
 
 S0 3 — S0 2 S— S0 2 C1— S0 2 I— 
 
 C0 2 — CO.C1— C6.A&. 
 
 s 2 6 5 -s 2 o 3 .ci 2 
 
 N0 5 — N0 8 .C1 2 
 
 Cr0 2 .Cl— Mo0 2/ CL W0 3 .C1. 
 
 Under certain circumstances, bodies belonging to different 
 groups are capable of replacing one another in combination 
 
COMPOUNDS. 
 
 53 
 
 without altering either the molecular type or the general rela- 
 tions of the primitive. Such derivatives are of course to be 
 classified with their primitives. Example. 
 
 Acetic acid, C 4 H 3 3 +HO. 
 
 Chloracetic acid, C 4 Cl 3 3 + HO. 
 
 5. Finally, the similar amphides of different groups may 
 be so nearly related in chemical properties, as to be placed 
 under the same natural group or sub-group of compounds. 
 Examples. 
 
 The protoxides of groups 3d and 4th, viz : 
 
 ZnO, MgO, HO. CdO, FeO, MnO, CoO, NiO, CrO. 
 The sesquioxides of groups 4th and 5th, viz : 
 Fe 3 O a ,Mn a 03,Co 3 03,Ni 3 03,Cr 3 0„Al a O„Cl 2 3 ,Zr a O s . 
 The bichlorides of groups 7th, 9th and 10th, viz : 
 PdCl 3 , PtCl 3 , IrCl 3 , 0sCl 3 , TiCl 3 , SnCl 3 . 
 The teroxides of groups 1st, 4th and 6th, viz : 
 
 S0 3 , Se0 3 , Te0 3 , Cr0 3 , Mn0 3 , V0 3 , W0 3 , Mo0 3 . 
 Certain oxides of groups 1st and 4th, viz : 
 
 S„0-, CIO., Mn o 0,—C10 7 , Mn o 7 . 
 
 6. Double compounds, that is to say, bodies composed of 
 two others, each of which is itself a compound, are classified 
 in a manner analogous to that just described for single com- 
 pounds, the classification depending mainly upon the electro- 
 negative or chlorous compound constituent. 
 
 Those double compounds whose constituents are members 
 of the same natural groups of single compounds are classified 
 together as a natural group of salts, and this group of salts is 
 again subdivided into eight sub-groups, comprising 
 
 1. Oxy-salts, 5. Chloro-salts, 
 
 2. Sulphur-salts, 6. Iodo-salts, 
 
 3. Selenio-salts, 7. Bromo-salts, 
 
 4. Telluri-salts, 8. Fluo-salts, 
 
 wherein the term salt is applied exclusively to the compounds 
 of bodies relatively acid with bodies relatively basic. For 
 example, 
 
 5* 
 
54 CHEMICAL CLASSIFICATION. 
 
 S0 3 , SeO„ Te0 3 , Cr0 3 , Mn0 3 , Mo0 3 , W0 3 , 
 
 beftmging to the same sub-group of acids, and 
 
 ZnO, MgO, CdO, HO, FeO, MnO, CoO, NaO, CrO, 
 
 belong to the same sub-group of bases, consequently the sim- 
 ilarly constituted compounds of these acids and bases form to- 
 gether a natural sub-group of salts, according to the princi- 
 ples which we have adopted. The 63 salts which would thus 
 be formed, are subdivided into 7 genera, viz : sulphates, sele- 
 niates, telluriates, chromates, manganates, molybdates, and 
 tungstates. 
 
 As the sulphates alone have been carefully studied, we will 
 cite from them a few examples of saline genera. 
 S0 3 +ZnO + 7HO $0 3 +CoO + 7HO S0 3 +MnO + 6HO 
 S0 3 ° + MgO+7HO S0 3 +NaO + 7HO S0 3 + MnO+4HO 
 S0 3 + FeO + 7HO S0 3 + CoO + 6HO S0 3 +FeO-f4HO 
 S0 3 +MnO + 7HO S0 3 +NaO + 6HO S0 3 + CdO+4HO 
 In these instances the sulphates are all hydrated, and ought 
 therefore to be regarded as triple compounds. Their formulae 
 however suffice to exhibit their analogies and the propriety of 
 placing them in the same genus. The peculiarity of their 
 constitution in containing an atom of constitutional water has 
 already been alluded to. 
 
 As further examples of natural sub-groups of salts we may 
 
 cite 
 
 PtCl +KCl OsCl„+KCl SnCl 2 + KCl 
 
 IrCl~+KCl PdcC+KCl TicC+KCl 
 which series of salts is arranged according to base, in conse- 
 quence of our imperfect knowledge of the other acids belong- 
 ing to the same group with the chloracids in the formulas. 
 The individuals of this genus are isomorphous. 
 
 The sulphur-salts are for the most part analogous to the 
 oxy-salts containing the same radicals. As these two classes 
 of salts have been more extensively and carefully studied 
 than any others, it would be easy to fill many pages with the 
 formulae of the sub-groups which they form. We shall con- 
 tent ourselves for the present by citing the formulae of a few 
 sulphur-salts. 
 
COMPOUNDS. 55 
 
 AsS 5 +NaS .AsS 5 +2NaS. r AsS 5 +3NaS. rSbS 5 + 3BaS„ 
 AsS 5 +KS .AsS 5 +2KS ^SbS 5 + 3NaS. |sbS 5 +3AgS 
 CS 2 +CaS. AsS 3 +RS. AsS 3 +2RS. AsS 3 +3RS 
 
 CS 2 + SrS. SbS 3 +RS. SbS 3 +2RS. SbS 3 +3RS 
 
 CS 2 + BaS. MoS 3 -fKS.— WS 3 + KS. rKS + HS 
 
 CS 2 + PbS. MoS 3 + NaS— WS 3 + NaS ^NaS + HS 
 
 Cgs 2 +AgS. MoS+LS— WS S + LS. cBaS+HS 
 
 £SrS+HS 
 7. The classification of the double salts will obviously de- 
 pend upon that of the salts of which they are composed, so 
 that we arrange in the same group, all double salts whose 
 proximate constituents belong to the same saline genera. As 
 examples of this mode we may cite the group of salts known 
 as the alums, whose formulas are 
 
 KO.SO3-fAUO3.3SO3 + 24aq 
 KO.SOa + Mn 2 3 .3S0 3 +24aq 
 KO.SO 3 + Fe 2 6 3 .3S0 3 + 24aq 
 KO.SO 3 + Cr~ 3 .3S0 3 + 24aq 
 Again, we have an exceedingly well defined group of double 
 salts in certain oxalates, represented by 
 
 3(K0.C 2 3 ) + Cr 2 O 3 .3C 2 3 +6aq 
 3(K0.G 2 3 )+Fe 2 3 .3C 2 3 + 6aq 
 3(KO.C 2 3 ) + Al 2 5 .3C 2 3 + 6aq 
 and in the well known double sulphates, 
 
 ZnO.S0 3 +KO.SO3 +6aq MnO.S0 3 +KO.S0 3 +6aq 
 MgO.S0 3 + KO.SO3 + 6aq CoO.S0 3 + KO.S0 3 + 6aq 
 CdO.S0 3 +KO.S0 3 +6aq NiO.S0 3 +KO.SO s +6aq 
 FeO.S0 3 +KOS0 3 +6aq CrO.S0 3 + KO.SO 3 +6aq 
 
 GnO.S0 3 + KO.SO 3 +6aq. 
 No double compounds of the other saline sub-groups are 
 known to exist, though there is every probability that double 
 sulphur-salts, chloro-salts and so on, will hereafter be described. 
 We are acquainted with two very remarkable double salts, one 
 oL whose constituents is an oxy-salt and the other a sulphur- 
 salt, they are represented by 
 
56 CHEMICAL CLASSIFICATION. 
 
 (WS 3 +KS)4-(N0 5 +KO) 
 
 ^ (MoS 5 +KS)-f(N0 5 + KO) 
 
 8. The principles which we have laid down for the classifi- 
 €ation of single and double salts, apply also to all compounds 
 consisting of three or more oxides, sulphides, chlorides, &c. as 
 well to those which consist of oxides united to chlorides, 
 iodides or other amphides. Under this last head all hydrated 
 salts ought in strictness to be included, together with hydrated 
 amphides, in which the water appears to play a different part 
 from that either of an acid or a base. The same remark ap- 
 plies to those compounds which contain excess of oxide, chlo- 
 ride or amphide in general, beyond and above what is neces- 
 sary to constitute a neutral salt with the particular acid or base 
 in the compounds. The oxides of zinc, copper and lead, and 
 the proto-chloride of mercury are peculiarly apt to form com- 
 pounds of this character, and the compounds thus formed we 
 regard as precisely analogous to ordinary hydrates. 
 
 To finish the task which we imposed upon ourselves in the 
 commencement of our essay, it remains for us to consider 
 briefly the analogies which subsist between simple and com- 
 pound bodies. 
 
 The discovery of Cyanogen, by Gay-Lussac, furnished the 
 first instance of a compound body which in all its chemical 
 relations bore the most exact analogy to an elementary sub- 
 stance. And since the period of its discovery the researches 
 of chemists have brought to light many other substances of 
 greater or less simplicity of constitution, which in the same 
 manner present all the chemical properties and relations of 
 true elements. It becomes then to us a matter of some im- 
 portance to determine the principles upon which these bodies 
 are to be classified and arranged according to the principles of 
 a natural system, since it is clear that they must occupy, in 
 any classification, a position essentially different from that of 
 an ordinary compound of two or more elements. If we con- 
 sider for a moment the properties of these compound elements, 
 if, without an Hibernicism, we may use such an expression, 
 
COMPOUNDS. 
 
 57 
 
 we shall find that, for the purposes of classification, they may 
 be divided into two classes, depending upon their analogies 
 with the amphigen group of our division, or with the non- 
 amphigen groups. Every compound radical is either electro- 
 negative or electro-positive, chlorous or zincous, an analogue 
 of zinc, or a former of acids and bases like oxygen and chlo- 
 rine. As a chlorous radical we may cite cyanogen ; as a zin- 
 cous radical, ammonium, and as a binary compound of a chlo- 
 rous with a zincous compound radical exactly analogous to 
 chloride of potassium we may cite cyanide of ammonium. 
 Show us a metallic chloride of a given molecular type and 
 we will produce a corresponding cyanide. Show us a salt of 
 potassium and we will place by its side a similar and even an 
 isomorphous salt of ammonium. It is upon these analogies 
 then that we base our classification of compound radicals. 
 Those which like oxygen or chlorine are capable of forming 
 acid or basic compounds with other radicals, we associate in 
 the same class with oxygen and chlorine, while those on the 
 contrary, which are most distinctly acidifiable or basifiable 
 like metals, we associate either in groups by themselves, ac- 
 cording to their analogies with each other, or in the same 
 groups with the simple substances which they may most 
 strongly resemble. It is of course to be constantly borne in 
 mind that compounds like simple radicals are always rela- 
 tively chlorous and zincous, and that the same substance may 
 at one time act the part of an electro-negative or chlorous rad- 
 ical, like chlorine in a metallic chloride, and at another time 
 that of an electro-positive or zincous radical, like chlorine in 
 chloric acid. Furthermore, we may remark in this connexion, 
 that the position of an element when acting electro-negatively 
 may be very different from the position which the same ele- 
 ment occupies when acting electro-positively or zincously. 
 For example, electro-negative chlorine is associated by the 
 strongest analogies with oxygen, sulphur, &c. while electro- 
 positive or zincous chlorine is isomorphous in combination 
 
58 CHEMICAL CLASSIFICATION. 
 
 with two equivalents of manganese (or sulphur, though this 
 last is rather inferred than demonstrated.) 
 
 wigain, chlorous sulphur is analogous to oxygen and the 
 other bodies of the amphigen group, whereas zincous sulphur, 
 as in sulphuric acid, is isomorphous with the elements belong- 
 ing to the fourth or manganese group, one equivalent repla- 
 cing one equivalent of metal. Thus the chlorates and hyper- 
 chlorates are isomorphous with the corresponding hypo-man- 
 ganates and hyper-manganates C10 5 +RO with Mn 2 5 + R0 
 ClO^ + RO with Mn 2 7 +R0, while the sulphates andseleni- 
 ates are isomorphous with the manganates and chromates, SO 3 
 + RO and SeO 3 + RO with MnO 3 + RO and CrO 3 + RO. The 
 relative states in which bodies exist in combination have not suf- 
 ficiently attracted the attention of chemists, and it is possible that 
 it is in these conditions that we are to look for the explanation 
 of many phenomena which are at present involved in obscurity. 
 Thus then, briefly and imperfectly, we have traced through 
 the different elementary bodies the prevalence of the law of 
 grades. We have endeavoured to show that the chemical 
 properties and relations of all substances differ only in degree, 
 and that they are therefore associated together in natural fam- 
 ilies or groups according to the different degrees of analogies 
 which subsist between them. We have already stated that it 
 is to Graham we owe the first suggestion of a natural 
 system of classification. We regard that idea as one of the 
 most important in the whole history of chemical science, and 
 as likely to lead us to deeper and clearer views upon almost 
 every point in theoretic chemistry, than we have yet been able 
 to attain. A just and comprehensive system of classification 
 has ever been a most powerful instrument in the advancement 
 of a science. As our knowledge advances, many changes will 
 be made in the groups of elements at present adopted as pri- 
 mary, many chasms will be filled up, many new points of 
 analogy discovered, many superficial resemblances detected 
 and exposed, but the principles and the main features of the 
 natural system of arrangement will, we most firmly believe, 
 
COMPOUNDS. 59 
 
 remain unchanged. And as with every new discovery, we 
 obtain a farther and clearer insight into the mysteries of na- 
 ture, we believe that the law of grades will become more and 
 more apparent, till at last it shall be acknowledged that in the 
 Inorganic as well as in the Organic kingdoms, Natura non 
 facit saitum. 
 
ilf 
 
 
COLUMBIA UNIVERSITY LIBRARIES 
 
 This book is due on the date indicated below, or at the 
 expiration of a definite period after the date of borrowing, as 
 provided by the rules of the Library or by special arrange- 
 ment with the Librarian in charge. 
 
 DATE BORROWED 
 
 DATE DUE 
 
 DATE BORROWED 
 
 DATE DUE 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 C28(l14l)M100 
 
 
 
 
— = 
 
 < i — g o 
 
 a 
 
 ■^ 
 
 |o G35 < 
 
 trtation on a 
 leiaical olassi 
 
 M-QD<#>7 ~ Gss