December 22, 1910] 



NATURE 



is one who advocates ""an emancipation from the en- 

 slavement of the slaughterhouse, with all its evils"; 

 and that he holds the view "that the butcher>- of 

 animals exists to-day solely because we demand the 

 victims. We have inherited the custom from bar- 

 barism ; there can be no doubt but that it will be 

 discarded altogether by the higher civilisation of the 

 future race." 



THE CHEMICAL ASALYSIS OF IRON AND 

 STEEL. 

 Die Untersuchiitigs-Methoden des Eiseus tmd Stahls. 

 By Dr. A. Riidisule. Pp. 395 + xvi plates. (Bern: 

 Max Dreschel, 1910.) Price 11 marks. 



THE title of this book is somewhat misleading. 

 It suggests a survey of the various methods 

 used in the experimental investigation of iron and 

 steel. In reality it is confined to one, viz., the 

 chemical analysis of the constituents. 



In his preface the author states that the analysis 

 of iron and steel is one of the most difficult problems 

 in analvtical chemistrv', and that his object has been 

 to give a critical summary of the methods used in 

 estimating each constituent, with a view to indicating 

 which are the most trustworthy. His book certainly 

 gives a singularlv complete account of the details of 

 the methods which have been employed in the last 

 fiftv vears. But it is only here and there that any 

 attempt is made to compare the limits of accuracy 

 of comparable processes. .\ technical chemist wishing 

 to estimate, say, phosphorus in a given steel to a 

 certain degree of accuracy, would have considerable 

 difficulty in making his choice from this book. Rapid 

 methods are now the order of the day, and the works 

 chemist has to adopt the quickest, consistent with the 

 necessar>' degree of accuracy. It is a pit}- that the 

 author has not borne this in mind more than he 

 appears to have done. 



.■\s is only to be expected, by far the greater part 

 of the book deals with the estimation of carbon, 

 manganese, sulphur, phosphorus, and silicon, elements 

 the percentage of which is demanded in all specifica- 

 tions where chemical analysis is included. In view 

 of the detail in which carbon is treated it is surpris- 

 ing to find that no differentiation is attempted between 

 ■'carbide carbon" and '"hardening carbon" in the 

 estimation of "combined carbon," although this is a 

 point of considerable importance. Moreover, there is 

 no reference to the so-called "missing carbon." 



The "alloy" steels containing special elements, e.g. 

 chromium, tungsten, molybdenum, nickel, vanadium, 

 &c., are well treated, and complete methods are given 

 for various combinations likely to be met in technical 

 practice. There is, however, no mention of uranium, 

 a modern constituent in the armour-plates of more 

 than one nav}'. 



It is significant of the neglect of the influence of 

 gases in iron and steel that the sur\'ey of all the 

 methods of estimation occupies only nine in a book 

 oT 380 pages. For oxygen only one trustworthy 

 method appears to be known. Nitrc^en has fared 

 somewhat better. There is only a scant}- reference 

 to carbon monoxide and none to hvdrogen and carbon 

 dioxide. Modern metallographical research is insist- 

 NO. 2147, VOL. 85] 



ing on the importance of the effect, particularly of 

 oxygen and nitrogen, on the properties of iron and 

 its alloys. For many years the pernicious influence 

 of phosphorus has been recognised. It is now coming 

 to be admitted that nitrogen may be anything be- 

 tween five and ten times as harmful. As yet no 

 specifications require the estimation of the gaseous 

 constituents in steel. The day is probably not far 

 distant when this will be demanded. 



The microscopic investigation of steels has led to 

 another demand, viz., for a knowledge of how the 

 various constituents exist in the metal. The present 

 method of returning them as though they existed as 

 such is entirely misleading. In the majority 

 of cases th?y are combined. Some are segre- 

 gated as special constituents ; others are dis- 

 solved in the main bulk of the metal. It is just 

 here that the present methods of chemical analysis 

 are woefully weak; indeed, the researches of Carnot 

 and Goutal stand almost alone as a praiseworthy 

 attempt to obtain infcM-mation on these matters. If 

 the author, when he comes to prepare a second edition 

 of his book, will include a chapter dealing with the 

 methods that have thrown light on the chemical con- 

 stitution as distinct from the composition of steels this 

 will add considerably to the value of an already useful 

 and trustworthy handbook. 



H. C. H. Carpenter. 



THE PSYCHOLOGY OF SCIENTIFIC INQUIRY. 

 Erkenntnistheoretische Grundziige der Naturwissen- 



schaften and ihre Beziehungen zum Geistesleben der 



Gegenivart. By Paul \'olkmann. Pp. xxiii4-454. 



Second, completely revised, and enlarged edition. 



(Leipzig and Berlin : B. G. Teubner, 19 10.) Price 



6 marks. 



THE second edition of this work (originally pub- 

 lished in 1896) appears as the ninth volume of 

 " Wissenschaft und Hypothese," a series which takes 

 its name from its first number — a translation of Poin- 

 care's w-ell-known essay. The author has adopted, 

 both in the subject-matter and the mode of exposi- 

 tion, numerous changes that are intended to fit the 

 book for its new r6le. In particular he has sought, 

 by avoiding unnecessary technicalities and by the 

 multiplication of examples from the histor}- of science, 

 to make his work useful and interesting to the lay- 

 man. There is no doubt that he has succeeded. The 

 book in its present form, though not comparable in 

 brilliance or charm with the name-volume of the 

 series, srives on the whole a sound and lucid treat- 

 ment of the matters with which it deals. Its chief 

 weakness is a certain lack of architectural unit\- and 

 clearness of plan. 



The author's general problem is to exhibit the 

 development of science as a psychological rather than 

 a logical process, the result of continued reactior\ 

 between objective realit\- and investigating minds. 

 This reaction is conditioned by certain postulates, such 

 as the postulate of congruence between the logical 

 necessities of abstract thought and the phenomena 

 which express physical "law-." Also it follows uni- 

 versallv the same general course, described by the 

 terms induction and deduction, isolation and super- 



