128 ORGANISM AND MECHANISM 



great complexity, it is not unfair to recall some of the dif- 

 ficulties, that the supposed mechanism has to form in 

 fertilisation a working unity with another mechanism as 

 complex as itself; that it has thereafter to divide over and 

 over again; that a part is sometimes as good as a whole; 

 and so on. It is sometimes easy to get twins from one egg 

 by shaking the first two cleavage-cells apart, and even at 

 the four-cell stage of the lancelet's development the same 

 method may result in quadruplets. It almost seems as if we 

 here reached a Euclidean reductio ad dbsurdum, of a mecha- 

 nistic interpretation. 



The central problem of development is differentiation, 

 and the biological study of this is not more than incipient 

 We cannot even elucidate the fact that the two cells into 

 which a germ-cell divides are sometimes exactly alike and 

 sometimes distinctly different. Out of apparent simplicity 

 there gradually emerges obvious complexity. As Roux puts 

 it, there is a self-manifestation of intrinsic manifoldness. 

 What was a clear drop a few hours ago is now a manifest 

 organism with nerve and sense-organ, food-canal and muscle. 

 It is the most wonderful thing in the world. Sometimes the 

 whole scene has changed when we return to observation 

 after the interruption of an hour's lecture. Be it under- 

 stood that no theory explains it, but while the biological 

 interpretation may try with some success to bring devel- 

 opment into line with, say, the organism's characteristic self- 

 repairing activity, the mechanistic interpretation has not 

 yet begun its task. It is a mere impious opinion that devel- 

 opment will one day be described in terms of mechanics. 



Another prominent fact in development is its regulated- 

 ness or correlation. Driesch and others have directed at- 

 tention to the power the embryo often shows of righting it- 



