244 AN INTRODUCTION TO MODERN GENETICS 



And we find that several different lines of Titanotheres underwent 

 trend evolution, in which the horns became larger and larger.^ It is not 

 known whether this connection of heterogonic growth and trend evolu- 

 tion is mere coincidence or not. 



Not all trends concern heterogonic growth. A very beautiful example 

 may be taken from the Chalk Micrasters,^ which were Echinoderms 

 rather like the modern heart urchin. We find here several slow con- 

 tinuous processes of evolutionary change affecting many different parts 

 of the shell; the general body outline, the position of the mouth and 

 apical disc, the depth of the anterior groove and the cross-sectional 

 shape, granulation and suturing of the ambulacral grooves all change 

 slowly in definite directions. Many other examples could be given. One 

 of the most famous is that of the reduction of toes in the horse, pictures 

 of which will be found in most textbooks. But we have nowhere nearly 

 so complete a series of overlapping forms for any land animal, even so 

 wide-ranging and common as the horse, as we have for marine inverte- 

 brates, and it is the latter, therefore^ which provide the most critical 

 evidence. 



Two further facts about trends require noticing. In the first place, a 

 trend may continue so far that it appears to lead to the extinction of the 

 gens. The latest representatives of any of the Gryphaea gentes show 

 such a curvature of the umbones that the opening of the shell appears 

 mechanically difiicult, and it is plausible to suggest that this was a 

 reason for the dying out of the line. Similarly, the latest representative 

 of a Titanothere gens may have such unwieldy horns as to appear 

 definitely ill-adapted. This appearance can never be confirmed by 

 actual observation and must remain a hypo±esis. But if it is true it 

 presents a particular difficulty to any theory which explains evolution 

 as a result of natural selection; and it is of course quite tmreconcilable 

 with any theory of evolution in terms of the inheritance of adaptive 

 acquired characters. 



The second point is that we find examples of two or more related 

 stocks which start off on several different trends, and in such cases gens 

 A may progress rapidly along trend P and more slowly along trend Q, 

 while gens B moves slowly along P and rapidly along Q. The trends 

 have then a certain independence. We shall see that this independence 

 has been elevated into a general principle for dealing with the pro- 

 gramme evolution on a larger time scale. 



(b) Programme Evolution. — Programme evolution is most strikingly 

 ^ Cf. Robb 1935. 2 Ro^e jg^p 



