346 



NATURE 



[Julys, 1919 



after the wings have been singed in a charcoal fire, 

 used as an article of food by the aboriginals. These 

 moths sometimes invade the cities and crowd into 

 houses and stores for the sake of darkness. At Mel- 

 bourne, in a large sugar store, I have noticed 

 M. decumanus collect the moths and eat the bodies, 

 rejecting the wings. 



There came under my notice lately at Pennant 

 Hills, near Sydney, a case of a curious article of food 

 for a rat. A rat gained access to the laundry attached 

 to my house, and for some weeks it used to drag 

 pieces of common soap behind any shelter and devour 

 them. That the soap was really eaten was evident, 

 because no particles were left lying about. Ulti- 

 mately I succeeded in trapping the rat, which was a 

 half-grown male, M. decumanus. An empty spring 

 trap was placed open in a box having an opening just 

 over the jaws. A piece of tissue-paper was arranged 

 over the jaws and the whole covered with a thin layer 

 of bran, a bait being laid at the far end of the box. 

 On examination I found the intestines empty and the 

 stomach gorged with fresh bran, which the rat had 

 •scooped up before entering the trap. Although I 

 searched carefully I could never find any means of 

 exit from the laundry or see the rat, but I presume 

 it must have got other food somewhere, for abso- 

 lutelv nothing edible was ever placed in the laundry. 

 The rats' excreta were always quite normal. 



Thos. Steel. 



Sydney, April 28. 



SOME RECENT ATOMIC WEIGHT DETER- 

 MINATIONS. 



THE story, adequately told, of the evolution 

 of ideas and the development of knowledge 

 concerning the stoichiometrical constants we term 

 atomic weights forms a most interesting chapter 

 in the history of the philosophy of chemistry. In 

 point of time it would extend over no very long 

 span. There are men living who are personally 

 cognisant of its most important phases, and some 

 of them in early life were acquainted with others 

 who may be said to have connected their own 

 epoch with that of those who witnessed the begin- 

 ning of experimental efforts to obtain quantitative 

 estimations of their values. 



The formulation of the laws of chemical 

 combination involved the necessity for exact 

 knowledge of the relative weights with which 

 substances enter into such combination, and, as 

 is well known, Dalton himself made tentative 

 trials to obtain some definite conception of their 

 measure. But Dalton was not a particularly 

 skilful or accurate experimenter; his apparatus 

 and methods of quantitative work were very crude 

 and even below the standard of his time. This 

 was fully recognised by his contemporaries, par- 

 ticularly by Berzelius, who may be said to have 

 been the first to attempt precise determinations 

 of atomic weights. The work of Berzelius and 

 his coadjutors marks, in fact, an epoch in the 

 history of the subject. 



Of course, as is now well understood, the germ 

 of Dalton 's ideas, although he probably was un- 

 conscious of it, is to be found in the work of his 

 predecessors, but it does not seem to be generally 

 known that Cavendish, in effect, postulated and 

 NO. 2592, VOL. 103] 



put into practice the fundamental conceptions ex- 

 pressed in the laws of constant, multiple, and 

 reciprocal proportions. He appears to have con- 

 vinced himself years before the time of Proust and 

 Berthollet that the same substance is invariably 

 composed of the same elements united in the same 

 proportion, and, as can be shown from his pub- 

 lished writings, he made quantitative analyses on 

 the implicit assumption of the other laws. This 

 was first pointed out by George Wilson, and has 

 been more fully developed in the course of a 

 critical examination of Cavendish's memoirs in 

 the Phil. Trans, for 1786 and 1788 on "Freezing 

 Mixtures," contained in an annotated edition of 

 his complete papers, published and unpublished, 

 which it is to be hoped the Cambridge University 

 Press may soon be in a position to issue. 



It would occupy more space than is available 

 to attempt to trace the several phases, which, 

 like milestones, mark successive stages in the 

 progress and development of knowledge concern- 

 ing atomic weights, nor is it necessary to set out 

 in detail the various reasons which have led 

 chemists to recognise the imperative necessity of 

 knowing these constants with the highest attain- 

 able precision. Philosophers like Berzelius always 

 desired the utmost accuracy in the abstract 

 interests of truth. But, to begin with,- the only 

 practical use of atomic weights, or combining 

 proportions as they were called by Davy, was in 

 quantitative efforts to elucidate the chemical com- 

 position of substances, and, considering the im- 

 perfections of quantitative methods, an approxima- 

 tion to exactitude suflficed. When substances 

 began to be bought and sold on the results of 

 analysis, atomic weights became of importance 

 in commercial transactions, but even then, for 

 the purpose of trade, no very high degree of 

 accuracy was required. Even the numbers of 

 Berzelius 's time sufficed for the determination of 

 exact formulse, and enabled the nature and pro- 

 gress of a chemical change to be traced with 

 precision. 



But in recent time, and with the development 

 of chemical theory, atomic weights have acquired 

 a wider importance and a new significance, and 

 a much higher degree of accuracy is demanded. 

 It is, in fact, almost useless to discuss certain 

 questions unless these constants have been rigo- 

 rously determined. Very much now depends upon 

 little differences — the little difference, indeed, fre- 

 quently makes all the difference. But, unless this 

 is established with reasonable certainty, it is a 

 waste of time to base an argument upon it. We 

 thus enter upon another and the latest phase in 

 the development of the subject. 



For this new departure, which may be said to 

 start with Stas, the chemical world is greatly 

 indebted to American chemists, such as J. P. 

 Cooke and his colleagues, Oliver Huntington and 

 Theodore Williams ; and to J. W. Mallet, Morley, 

 and Noyes. Prof. Theodore Williams has worthily 

 maintained the traditions of the Harvard school, 

 and it is largely to his work and example that the 

 present high standard has been reached. We 



