THIi GROWTH OF GENES 



symmetry and five stamens to the flower. This looks an even more 

 constructive change, although it is definitely backward to the simpler 

 ancestral type of Vcrbascum. 



Two things, however, need to be said about these mutations. 

 One is that the changes, whether backward or forward in evolution, 

 may nevertheless be physiologically negative. The other is that each 

 mutation by itself leads to a less satisfactory total system: it requires 

 other subordinate mutations to make it work well. We must ask 

 how well-adjusted is meristematic behaviour in Bergcff 's liverwort. 

 And how well adjusted is the system to which the sporangial 

 spermatozoids give rise in the fern ; The answers arc, very poorly. 



Thus these mutations do not affect the general rule. The steady 

 advance of evolution through a series of types, each adjusted and 

 indeed better adjusted than the predecessors which it ousts, cannot 

 proceed by large jumps no matter how potentially constructive these 

 might be. Rather it must proceed (as Darwin imagined) slowly, 

 by a number of small steps, accumulating until in total they give 

 the big changes. But once an advance has been achieved it may 

 be lost by a large jump backwards to some ancestral, and perhaps 

 more flexible, type as in the Antirrhinum. The adjustment of the 

 new type is not a new one in such a case, but is achieved by taking 

 advantage of gene complexes adjusted by selection long ago, though 

 perhaps more recently put to other work. In a word, it is possible 

 to go dov^oihill by big steps, but to go uphill only by small ones. 



How then, we must ask, do the genes, whose mutations produce 

 such drastic changes, come into being ? Now, as we saw, most of 

 the changes that we use in studies of inheritance represent decreases 

 or defects in the activities of the gene. These changes imply that 

 the gene itself is something built up and hence capable of degrada- 

 tion. They imply that other changes must also occur having the 

 effect of building genes up. And they imply that genes of different 

 degrees of building up will occur within the same organism. 

 Otherwise we should have to take the view (which was formerly 

 taken) that evolution is a mere unpacking of original units, which 

 themselves alone are specially created and free from the consequences 

 of evolution. 



The begiimings of any gene must consist of the simplest structure 

 which is capable (with tlie help of nucleic acid) of reproducing 



32S 



