606 



itCIENCE. 



[N. S. Vol. XI. No. 274. 



prevent the general use of Dr. MacDougal's 



^°°^- Charles B. Bessey. 



The Univeesity of Nebraska. 



De praMisohe ioepassing van Stoomschuif- en 

 Schaarbewegingen bij Stationaire, Locomo- 

 biel-, Locomotief- en Scheeps-machines door 

 C. Steuerwald ; Mit eene voorede van H. 

 A. Havener. By W. S. Auchincloss. 

 Leiden. A. W. Sijtboflf. 1899. Pp. 108. 

 Many illustrations. 



This book is a translation, into the Dutch, 

 of Auchincloss' well-known treatise on valve- 

 motion, of which a German version has long 

 been in type. The translator is a member of 

 the Faculty of the Polytechnic School of Delft ; 

 the introduction is written by Professor Ra- 

 venek of the same institution. There is no 

 lack of such works in the English, German and 

 French languages ; but the work of Auchincloss 

 excels in the simple and very clear manner in 

 which the graphical constructions are made, 

 "without preceding calculations, simply by 

 outlines on the drawing-board," as Professor 

 Ravenek says in his introduction. The treatise 

 is adjudged 'very suitable to be placed in the 

 hands of apprentices and draughtsmen ' as well 

 as of students in mechanical engineering. 



The British measures of the original are re- 

 placed in the translation by metric. 



This reproduction of the American work in 

 Dutch is one of the most gratifying testimonials 

 to the value of the work which has yet ap- 

 peared. The book is unusually well-printed and 

 its illustrations are exceptionally well-made. 



R. H. T. 



Mesure des temperature elevees. Par. H. Le 

 Chatelier et O. Boudouard. Paris, G. 

 Carre et C. Naud. 1900. Pp. 1-220. 

 In these few pages Le Chatelier and his as- 

 sistant have given a terse and useful account 

 of the principal methods of cotemporaneous 

 pyrometry. Measurement of high tempera- 

 ture has, as a rule, referred to the comparison 

 of different temperature functions, and the re- 

 sults obtained have therefore differed enor- 

 mously. The confusion has gradually subsided 

 however, in proportion as the air thermometry 

 of high temperatures has been more fully 



mastered. Le Chatelier makes a judicious 

 selection of standard temperatures in the in- 

 troductory chapters of his book and estimates 

 the probable error to be 1° between 200° and 

 500°, 5° between 500° and 800°, 10° between 

 800° and 1100° and upwards 50° above 1100°. 

 In the list of pyrometers which follows I 

 should have referred the calorimetric pyrometer 

 to Pouillet and perhaps included the viscosity 

 pyrometer. 



The brief account given of normal tempera- 

 tures as defined by Kelvin and their relation to 

 the air thermometer is intelligible, well digested 

 and practical in character, though these cor- 

 rections at high temperatures are of small 

 moment. An account is also given of the 

 standard (hydrogen) air thermometer of the 

 Bureau International at Sevres, which may be 

 taken as a preliminary model, since the nor- 

 mal air thermometer for high temperatures has 

 not yet been constructed. The authors might 

 have added that very definite steps are being 

 taken in this direction by Holborn and Day at 

 the Reichsanstalt. It has been shown that the 

 platinum-iridium alloy is impervious to nitrogen 

 rigid up to the highest industrial temperatures. 

 Nothing now stands in the way to prevent 

 high ^temperature measurements from attain- 

 ing the full precision of low temperature 

 measurements. 



The errors usually encountered in high tem- 

 perature thermometry make up Chapter III. 

 of the book, after which various historical 

 pyrometers are described from figures, and 

 critically discussed. It is interesting to note 

 that the errors of Pouillet were largely due to 

 the high value of the coefficient of expansion 

 then in vogue. Among the whole series the 

 interferential pyrometer of D. Berthelot may 

 be singled out as being peculiarly promising, 

 both on account of the simple and apparently 

 correct principle on which it is based, and on 

 account of its indefinitely high temperature 

 limit of application. 



In preference to platinum which is expensive 

 and iron which behaves anomalously, nickel 

 has been recently proposed for calorimetric 

 pyrometry. The authors give a series of appro- 

 priate data, and figures of available apparatus, 

 together with the probable inaccuracies of this 



