428 THE INSTABILITY OF THE HOMOGENEOUS. 



and by continuance of the process this membrane spreads 

 until it speedily invests the entire mass, as in mammals, or, 

 as in birds, stops short of that for some time. Here we have 

 two significant facts. The first is, that the primary unlike- 

 ness arises between the exterior and the interior. The sec- 

 ond is, that the change which thus initiates development, 

 does not take place simultaneously over the whole exterior; 

 but commences at one place, and gradually involves the 

 rest. Xow these facts are just those which might be inferred 

 from the instability of the homogeneous. The surface must, 

 more than any other part, become unlike the centre, because 

 it is most dissimilarly conditioned; and all parts of the sur- 

 face cannot simultaneously exhibit this differentiation, be- 

 cause they cannot be exposed to the incident forces with ab- 

 solute uniformity. One other general fact of like 

 implication remains. Whatever be the extent of this periph- 

 eral layer of cells, or blastoderm, as it is called, it presently 

 divides into two layers — the serous and mucous; or, as they 

 have been otherwise called, the ectoderm and the endoderm. 

 The first of these is formed from that portion of the layer 

 which lies in contact with surrounding agents; and the sec- 

 ond of them is formed from that portion of the layer which 

 lies in contact with the contained mass of yelk. That is to 

 say, after the primary differentiation, more or less extensive, 

 of surface from centre, the resulting superficial portion un- 

 dergoes a secondary differentiation into inner and outer 

 parts — a differentiation which is clearly of the same order 

 with the preceding, and answers to the next most marked 

 contrast of conditions. 



But, as already hinted, this principle, understood in the 

 simple form here presented, supplies no key to the detailed 

 phenomena of organic development. It fails entirely to ex- 

 plain generic and specific peculiarities; and indeed leaves us 

 equally in the dark respecting those more important dis- 

 tinctions by which families and orders are marked out. 

 Why two ova, similarly exposed in the same pool, should 



