1899] THE SCOPE OF NATURAL SELECTION 19 1 



cell, again occupying an internal position in the group, enlarges at the 

 expense of the surrounding cells, and when it has attained a certain 

 size ruptures from the capsule which surrounded it, extrudes two 

 nuclear portions of its substance (polar bodies), and if one of the 

 smaller active cells comes into contact, and fuses with it, it will 

 commence a series of cell divisions accompanied by increasing growth, 

 and develop into an adult hydra similar to its parent. This sexual 

 mode of reproduction very rapidly supplants all other forms ; it is 

 probable, therefore, that there is some immediate advantage resulting 

 to the organisms which reproduce in this way rather than by budding. 

 The most obvious difference in these two methods is that there is a 

 great reduction of tissue material, much less being required for this 

 mode of development than the other ; it is therefore less expensive to 

 the parent organism. Apart from this there is the additional factor 

 that it would be the most suitable for development, if direct climatic 

 accommodation does not take place, owing to its being the best means 

 of obtaining the requisite amount of variability. This reduction must 

 presumably be largely quantitative and not qualitative, since we find 

 that under very dissimilar conditions a complex hydra can be formed, 

 provided portions of both ectoderm and entoderm are preserved. 



Now, where this sexual mode of reproduction arises, we have to 

 consider a new set of conditions ; we find that each individual 

 appears to go through a stage of development, maturity, and decay, 

 and that during maturity the reproductive power of the whole 

 organism is best developed. 



Perhaps one of the most striking facts associated with the higher 

 forms of life is that these three periods of growth, maturity, and 

 decay in the whole organism do not correspond in time to similar 

 periods in the several different parts of the organism in question. 

 This fact appears to be universal in its application ; how is it to be 

 explained ? Now, as I have already noted, the most marked differ- 

 ence between unicellular and multicellular reproduction consists in the 

 fact that the latter develops chiefly by a quantitative evolution from 

 a cell which is quantitatively undifferentiated, while the former 

 reproduce by splitting off a portion of their structure, so that in most 

 particulars, except size, the parent and the offspring are identical. 

 Now one of the peculiarities of development and growth in one of the 

 higher organisms is just this quantitative development, and we must 

 assume that the morphological element is present, for it is inconceiv- 

 able that actual differentiation of structure could arise without some 

 structural difference for its starting-point. We are bound therefore to 

 assume two positions as essential to development: (1) Some basis for 

 the differences that are found in individual development which must 

 be of a structural and not a physiological nature, whether we call 

 them gemmules, physiological or morphological units, biophors or 

 stirp ; (2) that development consists largely in a reduplication of 



