494 SCIENCE PROGRESS 



properties of four-space, by the modified quadratic differential form ds*, and this 

 is precisely an intrinsic change, i.e. such as cannot be transformed away by 

 co-ordinate transformation (and, ne doubt, the author knows it well). There are 

 also some obscurities on page 25 in connection with "kinds of space not occurring 

 in Nature." But these are rather harmless. With the exception of these two 

 introductory chapters, the Report, which in its remainder is concerned almost 

 exclusively with the mathematical side of the subject, is carefully written and will 

 be found very useful as a handy substitute for the numerous and scattered original 

 memoirs of Einstein himself, of de Sitter, Droste, and others. We have here 

 a good presentation of the theory of Tensors, and of Einstein's field-equations, 

 integrated for the case of a gravitating particle (Chaps. III. and IV.). "The 

 Crucial Phenomena," treated in Chap. V., embrace the equations of motion of a 

 particle, illustrated by planetary motion, and leading to the famous formula for 

 the secular motion of the perihelion, the deflection of light and the shift of 

 spectral lines. The former "phenomenon" still awaits its verification, and the 

 latter has been rather disproved by St. John's (and more recently, by Evershed's) 

 observations of the Sun. Chap. VI. treats of the gravitation of a continuous 

 distribution of matter and of the propagation of gravitation. The Principle of 

 Least Action is the leading idea of Chap. VII., which thus reviews the energetic 

 (stress-momentum-energy) side of the subject. Finally, Chap. VIII. treats of the 

 curvature of the four-dimensional world, a subject connected with some more 

 recent innovations introduced by Einstein and de Sitter into the original theory. 

 It is to be regretted that, on the general conceptual side, no account has been 

 taken of a deep-reaching paper by Kretschmann {Ann. der Physik., February 

 191 7), and, on the physical side, of an equally interesting paper by Kottler {ibid. 

 August 1918). We repeat, however, that the bulk of the Report is likely to be 

 very useful, and those interested in the subject will be grateful to Prof. Eddington 

 for undertaking this by no means easy task. 



L. SlLBERSTEIN. 



Methods of Measuring Temperature. By Ezer Griffiths, D.Sc. [Pp. xi + 

 176, with 81 illustrations.] (London: Charles Griffin & Co., Ltd., 1918. 

 Price 8>s. 6d. net.) 



This book, from the pen of one of the body of scientific workers in the National 

 Physical Laboratory, is a very welcome addition to the literature of Physical 

 Science, more especially as the subject of which it treats (thermometry) is one in 

 which a considerable body of the progress is due to Englishmen. 



Within the compass of its pages will be found very full accounts of the five 

 main methods by which temperature is, now measured — viz. the mercurial ther- 

 mometer, the resistance thermometer, the thermocouple, the full-radiation 

 pyrometer, and the optical pyrometer. The general physicist, whose knowledge 

 of these methods has been obtained from current text-books of Physics, will find 

 a considerable amount of the detail necessary for precision work in each method 

 collected here, and the bibliographies at the end of 6ach chapter give the reader 

 the opportunity for consulting the original sources for further information. 



What strikes the reader is the remarkable increase in accuracy of deter- 

 mination of high temperatures which has taken place in this century — an increase 

 which (as Principal E. H. Griffiths in an Introduction points out) is in no small 

 measure due to the work of Prof. Callendar, Principal Griffiths himself, and 

 some other English physicists. 



