SCIENCE. 



[N. S. Vol. VIII. No. 186. 



tliree postulates uamed. As to the first, he 

 says that it is actually affirmed, or at least 

 assumed, in many theories of general biol- 

 ogy, of which the following are illustrations : 



1. Niigeli, De Vries and many other bi- 

 ologists think it necessary to believe in sep- 

 arate, discrete and numerable hereditary 

 units which exist in a kind of symbiosis in 

 the organism. The qualities which are rep- 

 resented by these units may be morpholog- 

 ical, physiological, latent, alternative, stages 

 in development, etc.; in fact, every differ- 

 ence in the organism is to be distinguished 

 as a separate constituent element. 



2. This postulate is also closely associ- 

 ated with the cell theory, which has become 

 unduly important under its influence. " It 

 is possible to so insist on the multitude, 

 on the similarity and on the independence 

 of cells as to deny the supreme individual- 

 ity of the bodJ^ The whole organism, it is 

 said, is but a colon j^ of these, the true indi- 

 viduals, and the secret of its form is to be 

 found in their habits of growth, reproduc- 

 tion and differentiation. And so the ques- 

 tion of the whole and the parts is removed 

 from the sphere of the body, in which we 

 have some opportunity of studying it, only 

 to be repeated in the microscopic sphere of 

 the individual cells. * * * If the individ- 

 uality of the body is to be slurred over we 

 have a right to expect that some architect- 

 ural principle should be found in the cell 

 itself. * * But I find no such attempt to 

 fill up the conception of cells as anthropo- 

 morphic agents. * * We are not likely to 

 find within an individual abstracted from a 

 system in which it is onlj' an element, the 

 principle of the architecture of the whole 

 system. * * Myriads of miserable Egyp- 

 tians carried stones to the Pyramids ; but 

 no microscopic watching of any of these, 

 stone and all, would ever explain the Pyra- 

 mid itself." 



3. Another illustration of this first 

 postulate is found in the doctrine of the 



independence of parts, particularly put 

 forward by Eoux in his ' Struggle of the 

 Parts.' " There must be some curious 

 fascination about this conception of struggle 

 that it should be introduced into the ex- 

 planation of the parts of that which is the 

 most perfect and unique unity we know." 



4. The ordinary conception of independ- 

 ent variability of parts implies independ- 

 ence of qualities. If variations are really 

 independent then we may at once give up 

 the unity of the organism. The author 

 argues that there is no such thing as inde- 

 pendent variability, that all variations are 

 correlated. Darwin's cases of correlated 

 variations, viz., hairless dogs having im- 

 perfect teeth, white cats being deaf, etc., are 

 only whimsical instances of a general law 

 of correlation of parts. Natural selection, 

 by insisting on independent variability, is 

 unable to explain the numerous coordinated 

 variations necessary to make the variation 

 of a single part effective. The same diffi- 

 culty is met, though to a less extent, by the 

 advocates of use inheritance, for here also 

 qualities are considered as primarily in- 

 dependent. " Confusion inevitably awaits 

 anj^ theory which moves bj^ the disintegra- 

 tion of the individual into self-sufficient 

 and primarily unrelated parts." 



5. " This first postulate is further shown 

 in the ordinary biological treatment of 

 funcUonless parts, which are supposed to 

 exist in their own right and in virtue of a 

 separate inheritance. A functionless part 

 of an organism is not useless ; it is merely 

 useless in a certain manner. =i= * * Nothing 

 organic is functionless, except for a certain 

 special abstract point of view." Every 

 part is in some way related to every other 

 part, and the very fact that a structure ex- 

 ists at all is evidence that in the process of 

 its origin, development and maintenance it 

 is functionally related to other parts. " One 

 is apt to hastilj^ assume that the significance 

 01 ^h9 part to the individual has nothing to 



