THE AMERICAN NATURALIST 



[Vol. XL 



^' daily lectures preceding the laboratory work and describing the 

 -experiments to be performed." It is, nevertheless, remarkable for 

 its freedom from the pedantic, cut-and-dried, schedule method of 

 presentation which so frequently characterizes elementary laboratory 

 manuals, for many of Professor Sabine's pages are interesting reading 

 as such, and throughout, "too specific instruction" has been avoided 

 as tending "not only to deprive the student of initiative but also to 

 make any departure in the apparatus confusing." 



As a matter of fact the spirit of the book would have been better 

 expressed by reversing the order of these two clauses, for "in the 

 majority of cases the description is purposely not such as will admit 

 of a mechanical and unintelligent interpretation." In particular, 

 the three-page introduction is an unusually fine presentation of the 

 point of view from which a student should attack the work which is 

 to follow. 



The experiments described are representative of nearly the whole 

 range of elementary physics. They should properly be preceded 

 by the still more elementary work of a modern high-school course, 

 as much of the apparatus r(M|uires eoniparatively skilful and ap])re- 

 ciative liandliiiir. Two short appeiulices on " siu;inficai\t figures" 



BIOLOGY 



Clements's Research Methods in Ecology ^ is the outcome of some 

 eight years of practical work by the author in the experimental study 

 of thr faftoiN tiiat (letcnniiK^ rhe distribution and adaptive modifi- 

 catioiH of plaiit-^. Smdciit- (.t ihi> comparatively new branch of 



The author point-; out that the greater part of so called ecological 

 study ha> hitherto hcc, very >,ipcrficial and of coinparaliv.-ly little 



rately the .ev.-ral factor^ thai .leter.niue for ea<'h speci.-s its particular 



versity Publishing Co., 1905. 8vo, xvii + 334 pp!/85 figs. 



