November 25, 1920] 



NATURE 



401 



but those on Inorg^anic Chemistry show no sign of 

 any decline in popularity, and maintain their posi- 

 tion almost unchallenged as the standard work on 

 chemistry in the English language. It is not 

 easy at the first attempt to discover what are the 

 special qualities that give to Roscoe's book this 

 It tribute of perpetual youth and long-sustained 

 .tility, but the refusal of the author to sacrifice 

 either clearness of exposition or scholarly writing 

 in order to reduce the size of the treatise, or to 

 overcrowd its pages with detail, has perhaps been 

 one of the most important factors in securing 

 these enviable attributes. Thus it is still possible 

 to turn to the work for the detailed story of the 

 investigation of the fixed or variable oxygen-con- 

 tent of the atmosphere, or of the composition of 

 the distillate from hydrochloric acid, without find- 

 ing that the narrative has been so abbreviated as 

 to be valueless except as a guide to the original 

 papers. The editors of successive editions must 

 have exercised considerable restraint in order 

 to allow a full account to be preserved of 

 <'xperimcnts which were becoming too old to 

 be modern, but were still too modern to be 

 classical. 



In bringing out the new edition. Dr. Cain has 

 been handicapped by the fact that he has no longer 

 lifen able to refer his work to Roscoe himself for 

 .pproval; but, having Ijeen associated with Roscoe 

 in the preparation of the preceding edition, he has 

 had special advantages in striving to preserve the 

 character and style of the book, and has suc- 

 ( cdcd so well in his task that the later dates 

 ', liich now appear in the footnotes are the most 

 iinspicuous marks of modernity. Here and there 

 a paragraph remains which shows signs of obso- 

 h-scence, and in an occasional instance (e.^. the 

 electrical method of making carbon disulphidc) 

 modern work has escaped notice ; but the new 

 edition is a worthy successor to those that have 

 gone before, and will contribute its share to the 

 long life of the treatise. 



One fault which was formerly characteristic of 

 Roscoe's "Chemistry" has almost disappeared in 

 the new edition — namely, the conversion of classical 

 apparatus into a modern form, without any indica- 

 tion in the text of the transformation that had been 

 ' Ifected. Only one example of this curious pro- 

 rcss jippears to have survived — namely, the in- 

 troduction of a gas furnace with a row of Bunscn 

 burners in the figure illustrating the experiments 

 on the composition of air carried out by Dumas 

 and Koussingault in 1841, although the joints of 

 the apparatus arc still shown with the original 

 rubber bandages instead of rubber tubing. This 

 last link with an old tradition will perhaps be 

 NO. 2665, VOL. 106] 



broken when the time comes for a sixth edition 

 to appear. 



It is some satisfaction that the printing and 

 paper show no sign of deterioration, so that the 

 appearance of the book is «as attractive as in 

 former years. Roscoe's "Chemistry" has never 

 stooped to the use of black type as a means of 

 emphasis, nor to the use of smaller type for 

 matter of less importance. Even the conven- 

 tional division of the text into chapters is missing. 

 These features have given to the book a character 

 of its own, which clearly appeals to the more 

 scholarly type of reader, even if the student finds 

 that he is obliged to read the book instead of 

 skimming through it from one key-word to 

 another. The student will still find, however, that 

 he has in Roscoe the best available guide to the 

 literature of inorganic chemistry, directing his 

 attention to all the more important papers, and 

 passing lightly over the mass of detail which has 

 converted so many of the larger works from text- 

 books into dictionaries. Roscoe's "Chemistry," in 

 spite of its increasing size, still possesses all the 

 essential qualities of a book rather than of a cata- 

 logue, and this is perhaps the principal reason 

 why its approaching jubilee is unaccompanied by 

 any marks of old age. T. M. L. 



Archimedes. 

 Archimed.es. By Sir Thomas Heath. Pp. ii-(-38. 

 (Pioneers of Progress Series.) (London : 

 S.P.C.K. ; New York: The Macmillan Co., 

 1920.) Price 25. net. 



BY the general consent of all competent judges 

 Archimedes is one of the greatest mathe- 

 maticians the world has ever seen. It is not easy 

 to justify this opinion to a popular audience, most 

 members of which know little and care less about 

 mathematics; but Sir Thomas Heath's book ought 

 to succeed in making the ordinary reader under- 

 stand to some extent the nature of Archimedes' 

 discoveries, and in arousing interest in the 

 achievements of Greek mathematicians. 



Chap. i. gives such fragmentary (and often 

 legendary) notes as we have on .Archimedes' 

 personal career; chap. ii. is an excellent account 

 of Greek geometry before .'\rchimedcs' time; 

 chaps, iii.-vii. give analyses of Archimedes' ex- 

 tant works. Special attention may be directed to 

 the paragraphs (pp. 31-35) on the "Method," dis- 

 covered by J. L. Heibcrg so lately as 1906 in a 

 palimpsest at Constantinople. This work shows 

 how Archimedes was led to some of his theorems 

 by quasi-mechanical considerations. It should be 



