134 Notices respecting New Boohs. 



Now Regnault's experiments were conducted on a large 

 quantity of C0 2 quite removed from capillary influences, so 

 that there is no difficulty in understanding the reason for the 

 apparent conflict of his result with those of Andrews. The 

 true critical temperature of C0 2 was near 42°*5 C. ; Andrews 

 had found the critical temperature of C0 2 when enclosed in a 

 certain capillary tube to be 31°, a temperature which depends 

 on the tube. 



[To be continued.] 



XIY. Notices respecting New Books. 



Metretike : or, the Method of Measuring Probability and Utility. 

 By F. Y. Edgewoeth, M.A. (London, The Temple Company ; 

 6S pp.) 



TN this pamphlet Prof. Edge worth, as per his Table of Contents, 

 -*- discusses " The relation of the Calculus of Probabilities to 

 the Philosophy of Utility is partly (i) of Similarity (pp. i-52) ; 

 partly (ii) of Identity (pp. 52-56). 



" (I.) Similarity exists both (A) as to the methods of calculation 

 and (Bj as to the problems proposed. 



" (A) The similarity is contemplated under five headings, consti- 

 tuted each by the introduction of an attribute, which tends to render 

 the calculation inexact (pp. 1-50). Under each of these headings 

 it is argued that there have been granted to the Calculus of Proba- 

 bilities postulates analogous to those which the Calculation of Utility 

 requires to be granted. 



" (B) The canons of Belief and Conduct are of similar origin and 

 authority (pp. 50-52). 



" (IT.) The Theory of Errors, inasmuch as, philosophically, the 

 qucesitum is rather the ' most advantageous ' than the most probable 

 combination of the given observations, involves the principle of 

 Utility (pp. 52-56)." 



Then follow two appendices, " On the Method of Least Squares " 

 (pp. 57-66), and " A List of Cognate Writings, bv the same Author " 

 (p. 67). 



Many of these papers are already familiar, or, at any rate, known, 

 to our readers, as they have been printed in the ' Philosophical 

 Magazine ' at varying dates, from Oct. 1883 to quite a recent issue 

 in the present year ; others, whose titles are given, are on their way. 

 The Author touches on many themes, and it goes without saying 

 that there is much matter furnished for careful rending and weighing. 

 Appendix I. may be commended to our mathematical readers, as 

 containing an interesting account and discussion of the Method of 

 Least Squares. 



