April 17, 1914] 



SCIENCE 



581 



Life Histories; Environmental Change; Meta- 

 bolism, Growth, etc.; Adjustment between the 

 Environment and Animal. Its scientific value 

 lies in the author's outline for the organiza- 

 tion of the science. 



In the first chapter, the subject is divided 

 into " individual ecology " or the ecology 

 of individuals and species, and " aggre- 

 grate ecology," or the ecology of taxonomic 

 groups of species, genera, families, etc. These 

 two divisions have usually not been recognized 

 separately. The distinction is good, but the 

 two divisions taken together are coordinate 

 with his third division, " associational ecology," 

 or the ecology of communities. Chapters II.- 

 IV. are devoted to ecological surveys and 

 methods of conducting them. The author 

 rightly deplores the tendency of museums to 

 rate the work of collectors and expeditions on 

 the basis of number of specimens brought 

 back, as this discourages the recording of 

 ecological facts. The methods of collecting, 

 preserving and arranging notes and speci- 

 mens, and securing proper identification of the 

 latter are given. These chapters will be of 

 material aid to those undertaking field ecolog- 

 ical study. 



Although divided under several chapter 

 headings, the remainder of the book consists 

 essentially of about 90 pages of classified 

 bibliography. The references are of a 

 diversified type and are intended to guide 

 the worker to needed information, ranging from 

 the making up of the sometimes necessary 

 camping outfit, to the preparation of his re- 

 sults for the printer. They are classified under 

 general headings and many are followed by 

 statements as to contents. Representative 

 papers on the environment, animal communi- 

 ties, struggle for existence, physiology, behav- 

 ior and many other topics are included. The 

 references are complete enough to give valu- 

 able suggestions to workers from almost any 

 point of view in ecology. The comments on 

 the more important older ecological papers 

 make it clear that a considerable number of 

 incomplete attempts at organization of knowl- 

 edge of animal communities have been re- 

 corded. Thus by means of the references, the 



book gives the history of the development of 

 the science. V. E. Shelford 



University of Chicago 



SPECIAL ABTICLES 



THE " GOLDEN MEAN " IN THE INHERITANCE OF 

 SIZE 



In bulletin 242 of the New Jersey State 

 College Experiment Station, page 39, appears 

 the following statement: 



The size and shape of the Fi (tomato) fruits 

 are the geometric means between the size and 

 shape corresponding to those factors of the parents, 

 which were active in crossing. 



This statement was based upon the measure- 

 ments of many thousands of tomato fruits. 

 A mode of inheritance in which xCFJ = \/db 

 assumes that in the union of gametes repre- 

 senting factors for size a multiplication and 

 the extraction of a square root take place. I 

 was unable at the time to explain how that 

 might occur, and so far no attempt by any one 

 else to interpret the significance of my results 

 has come to my notice. In a forthcoming 

 bulletin I shall present details of Fa and Fa, 

 and this paper is published solely to set forth 

 the nature of the principle of size inheritance 

 by the " golden mean," to indicate its bearing 

 upon vital questions on subjects in heredity, 

 and to establish the priority of its discovery. 



When two factors for different sizes of 

 homologous parts meet in a cross, the re- 

 sulting F, size is commonly intermediate in 

 such parts as are not greatly subject to fluctua- 

 tion. If a and h represent the parental size 

 characters, it has been tacitly assumed by most 

 investigators that the Fj size is (a J7 ?')/2. 

 In tomato fruits I found it to be \/ab. On 

 the face of it the average (algebraic mean) 

 seems the more probable, but when used as the 

 basis for the comparable inheritance of lines, 

 surfaces, and volumes, it becomes impossible. 

 For example, let plants be crossed, which have 

 spherical fruits of diameters a and h, and let 

 us assume, for the sake of simplicity, that all 

 cells constituting the volumes of both fruits 

 are of equal dimensions. Then the ratio be- 

 tween parents and F^ of the number of cells 



