134 



NA TURE 



[June i i, 1891 



stars." This seems to be a mistake, as three fourth type 

 stars are found in the Draper Catalogue. They are :— 



Name R.A. Decl. PicVerins's Photo. Dune'r. 



o h. m. o - 'etier. mag. 



D.M. + 17-1973 - 849+1736... Y, - J;^^ ■■ Tvin 



D.M.+68-6I7 ... 10 38 + 67 56 ... A? ... 6-50 ... IV. 

 D.M. + 76734 •■• »9 25 + 76 22 ... E ... 708 .. iv. ... 



These stars each occur on one plate only. 



The photographs show that the following stars have 

 bright lines in their spectra :— 



Known variable stars : f Aurigae, a Orionis, C Gemin- 

 ■orum, a Herculis, /3 Pegasi. ^^ ^ • 



Suspected variable stars : a Cassiopeiae, 66 Ceti, p Per- 

 sei, a Tauri, S Canis, /3 Geminorum, a Bootis, /3 Ursae 

 Minoris, /3 Cygni, y Cephei. 



Other stars showing bright lines, not hitherto detected, 

 are : t Ceti, y Andromedoe, k Persei, a Persei, v Persei, 

 80 Tauri, \ Aurigae, f Cancri, <t^ Ursae Majoris, o Leonis,, 

 7 Leonis, | Ursae Majoris, 43 Comae, o Bootis, 7 Scorpii, 

 /3 Coronse, f Herculis, v Herculis, /x Herculis. 



T. E. ESPIN. 



SOME ASPECTS OF STAS'S WORK. 



FOR the last thirty years Stas's work has set the stand- 

 ard of excellence in all that relates to atomic weight 

 determination. The literature of the subject teems with 

 references to his classic memoirs, which have come 

 to be regarded by chemists in the light of canonical books. 

 Admiration of the almost magical accuracy of Stas's 

 results seems somewhat to have diverted attention from 

 the rare philosophical insight displayed in ih^ plan of his 

 researches. Yet it is not too much to say that, while we 

 owe the conception of the atomic theory to Dalton, Stas 

 first placed the theory on a sound experimental basis. 



It was in the year 1843 that Dumas and Stas's value for 

 the atomic weight of carbon recalled attention to the 

 hypothesis of Prout which had hitherto met with little 

 favour on the Continent. The subsequent work of Dumas 

 and of de Marignac led these chemists to support the 

 hypothesis in a modified form. In i860 appeared the first 

 series of Stas's researches, " Sur les Rapport reciproques 

 des Poids atomiques." In the introduction to his paper 

 the author stated his conviction that these researches 

 furnished proof, as complete as the nature of the subject 

 admitted, that the hypothesis of Prout was a pure delusion 

 — that there was, in fact, no common divisor between the 

 atomic weights of the elements. In reviewing the work 

 of Stas, de Marignac admitted the impossibility of recon- 

 ciling the concordant results obtained by Stas and himself 

 with even the modified form of Prout's hypothesis. Yet 

 he regarded the dictum quoted above as too absolute in 

 character. It was by no means established, he contended, 

 that the constituents even of stable compounds are present 

 exactly in the proportion of the atomic weights. De 

 Marignac's criticism struck at the very basis of the atomic 

 theory but this by no means deprived it of its weight. 

 The laws of chemical combination are the experimental 

 basis of the atomic theory, and Stas admitted that these 

 laws had never been proved as " lois mathdmatiques." 

 Writing in 1865, in the introduction to his " Nouvelles 

 Recherches," he remarks that some of the fundamental 

 ideas of chemistry, which are generally taken as having 

 been proved, are as a matter of fact far from being so. He 

 considers that the constancy of composition of chemical 

 compounds has been experimentally established, but points 

 out that this does not constitute a proof of the law of 

 -constant proportions, the law, viz., which states that the 

 particular proportions in which two elements are combin- 

 ed in a certain compound is a cotistant proportion in all 

 the compounds which contain those elements. This had 



NO. II 28, VOL. 44] 



never been proved, yet it was only in this way that the 

 position of the atomic weights as constants of nature could 

 be estabhshed. The so-called law of multiple proportions 

 Stas referred to as an hypothesis of Dalton, pointing out 

 that the very rough analyses on which Dalton relied — of 

 which the error is frequently more than 10 per cent. 

 — as well as the results obtained by WoUaston and 

 by Gay-Lussac, were at most capable of establish- 

 ing a "loi limits." The state of science at the 

 time demanded a thorough re-examination of the basis 

 of the atomic theory. Stas realized this need, and took 

 upon himself the burden of the task. The conception 

 and plan of the " Nouvelles Recherches sur les Lois 

 des Proportions Chimiques" show the mind of a 

 great thinker not less clearly than the results of the work 

 exhibit the skill of a master in the art of experiment. 

 The " Nouvelles Recherches" contains a verification as 

 "loi mathdmatique" of the law of conservation of mass, 

 in the complete synthesis of silver iodide, and the com- 

 plete analysis of silver iodate. The constancy of compo- 

 sition of chemical compounds was subjected to a crucial 

 test in the experiments on ammonium chloride, and the 

 constant proportion between the combining weights of 

 elements in different compounds was tested in the conver- 

 sion of silver iodate, bromate, and chlorate, to the corre- 

 sponding haloid salts. The law of equivalent proportions 

 was verified by the concordant results obtained for the 

 atomic weights of silver and of the alkali metal? deter- 

 mined as functions of those of iodine, of bromine, and of 

 chlorine respectively, oxygen forming the common stand- 

 ard. One cannot help regretting that the law of multiple 

 proportions was not also made the subject of investigation . 

 The most suitable examples occur among gaseous sub- 

 stances, and the operations of gas analysis were foreign to 

 the methods of manipulation employed by Stas. The 

 complete analysis of nitrous oxide was indeed contem- 

 plated in order to determine directly the atomic weight 

 of nitrogen as a function of that of oxygen, but the idea 

 wais abandoned owing to the difficulty of constructing the 

 necessary apparatus. 



The work on the laws of combination furnished fresh 

 materials for the examination of Prout's hypothesis. Stas's 

 comments on the origm of this hypothesis possess a high 

 degree of philosophic interest. The remarks to which we 

 more particularly refer are the following : — " Lorsqu'on 

 remonte a I'origine de I'hypoth^se (de Prout) on s'apergoit 

 immediatement qu'elie doit sa source a un pr^jug^ ou, si 

 Ton veut, a un opinion pr(^concue, concernant la simplicite 

 des lois de la nature. Pendant longtemps les chimistes 

 comme les physiciens, des I'instant qu'ils ont vu certains 

 faits se reproduire avec une apparetice de regularity, ont 

 cru a I'existence d'une loi naturelle susceptible d'etre 



exprimde par une relation mathdmatique simple 



C'est a cette tendance, d'ailleurs tres-naturelle, qu'on doit 

 I'hypothese de Prout." Dalton's enunciation of the law 

 of multiple proportions is relegated by Stas to the same 

 category as a generalization on insufficient data. 

 Mendeleefif has remarked (Faraday Lecture, 1889) that 

 the periodic law has shown that the masses of the atoms 

 'mcr&2i^Q per saltum, in a manner which " is clearly con- 

 nected in some way with Dalton's law of multiple propor- 

 tions.'' Dalton was more fortunate than Prout. The 

 , combining proportions are expressible by a simple 

 \ mathematical law, whilst the atomic weights are only to 

 be represented by a complicated formula which may have 

 some such form as that proposed by Carnelley. 



The " Nouvelles Recherches" appeared in 1865. The 

 first paper on the periodic system \\ as read before the Rus- 

 sian Chemical Society in the spring of 1869. It is curious 

 to reflect that the foundations of the atomic theory had 

 hardly been made sure by Stas ere they were called upon 

 to bear the magnificent structure raised by Mendeleefif. 



V. C. 



