March i i, 1920] 



NATURE 



31 



duced the motion, that he had in mind the 

 ordinary member of the public, and particularly 

 the child in the elementary school, and appeared 

 to be under the impression that the expert was 

 already sufficiently well cared for. An instructive 

 feature of the debate is the almost entire absence 

 of any reference to science in general, or to the 

 Natural History Museum, in particular. Another 

 revealing point is suggested by a passage in Lord 

 Crawford's reply for the Government, in which, 

 in reference to Lord Bryce's proposal that a central 

 scientific department of the Government should be 

 set up, he said : " Among the purposes for which 

 the Scientific and Industrial Research Department 

 was set up . . . is actually that of acting as a 

 central advisory body on any scientific question in 

 relation to any Government Department " ; for, 

 apropos of this statement, we must observe that, 

 whereas every administrative Department is repre- 

 sented by at least one assessor to the Advisory 

 Council, that Department appears to be unaware 

 of the existence of the Natural History Museum. 

 Lord Sudeley's motion was eventually by leave 

 withdrawn ; nevertheless, we hope that the matter 

 will not be allowed to rest there. We think, 

 indeed, the question of sufficient importance for 

 the consideration of a Royal Commission, the 

 terms of reference of which should include the 

 .system of remunerating the specialist, who at 

 present enjoys a much lower scale of salary than 

 the administrator of corresponding standing, and 

 we strongly urge the Government to appoint one 

 with the least possible delay. 



Mathematical Cosmogony. 



Problems of Cosmogony and Stellar Dynamics. 

 By J. H. Jeans. Being an essay to which the 

 Adams prize of the University of Cambridge 

 for the year 1917 was adjudged. Pp. viii + 

 293 + v plates. (Cambridge: At the University 

 Press, 1919.) Price 215. net. 



IN a well-developed science two branches are 

 broadly to be distinguished. In the one, an 

 existing state of things is investigated. The 

 subject of research is events immediately con- 

 nected, forms, functions, and the laws which 

 govern them. The other branch generally marks 

 a later stage, and, basing itself on the results of 

 the first, .seeks to reconstruct from the present as 

 complete a picture as possible of the past and 

 even of the future. As in the conception which 

 underlies the theory of relativity, the present, 

 which is the limited subject of experience, is 

 merely a section in time from which a higher 

 NO. 2628, VOL. 105] 



manifold is to be deduced. When the subject- 

 matter is biological, the outcome is a theory of 

 evolution. When it coincides with the domain of 

 astronomy, the result is more specificallv recog- 

 nised as a scheme of cosmogony. 



There are at least three methods by which 

 attempts have been made to formulate such a 

 scheme. The first, and most trivial, is to .seize on 

 some remarkable phenomenon, like Saturn's rings 

 or a spiral nebula, and to see in it a clue which 

 can be followed up more or less plausibly with 

 the help of an exuberant and unfettered imagina- 

 tion. Progress on that line is naturally as limited 

 as it is precarious. The second method is illus- 

 trated in its highest form by the work of Sir W. 

 Herschel. It is the way of comparison and classi- 

 fication. The Draper classification of stellar 

 spectra by Pickering is an apt modern example. 

 ^^■ ithout preconception, except such as readily van- 

 ished in the light of fuller experience, almost all 

 the stars fell into an ordered sequence, which 

 became more complete and continuous as the 

 material accumulated. To connect the ascertained 

 sequence with a time scale was natural. But the 

 problem has not proved quite so simple as at one 

 time it appeared. In general, when the process 

 is exceedingly slow and the section of experience 

 correspondingly thin, the very direction of the 

 scale is ambiguous, and the method requires to 

 be supplemented by some additional principle. .'\ 

 third method remains. This cons^ists in the study 

 of models having a definite specification as nearly 

 as possible in accordance with cosmic examples, 

 but always within the power of analysis to discuss. 

 The behaviour and development of such a model 

 are traced to their logical consequences with full 

 mathematical rigour, and only after this has been 

 done is an attempt made to find their counterparts 

 in the actual universe. This is the profoundly 

 difficult but promising method adopted by Sir 

 George Darwin, by Poincare, and by Mr. Jeans in 

 the work under notice. 



It is curious how great are the difficulties which 

 surround problems capable of the simplest state- 

 ment. Three balls are thrown in any given way 

 in empty space. .Ml the intractable difficulties of 

 the problem of three bodies are involved in discuss- 

 mg the subsequent motion under their mutual 

 attractions. Or again, to take the fundamental 

 problem of the present subject, a mass of liquid is 

 stirred into rotation and left to find its shape under 

 its own attraction. What figure will it assume 

 when isolated in space? The following quotation 

 from Thomson and Tait may be worth recalling : — 



" During the fifteen years which have passed 

 since the publication of our first edition, we have 

 never abandoned the problem of the equilibrium 



