PALEONTOLOGY, ITS AIMS AND METHOD 21; 



to the production of a definite final structure, and that it is 

 hence legitimate to speak of an evolutionary trend. 



2. That an evolutionary trend is irreversible. Once com- 

 mitted to a course of evolutionary change a stock must follow 

 it to the end, and as it does so its power of giving rise to 

 branches with diverse trends becomes more and more reduced. 

 In consequence of this loss of potentiality the greater groups, 

 which are really initially distinguished only by their different 

 trends, necessarily separate very soon after the establishment 

 of the group from which they spring. 



3. That in consequence of the imposition of a phyletic 

 trend allied stocks pursue parallel series of changes, the accuracy 

 of the parallel depending on the closeness of the relationship, 

 and that in consequence, as Prof. W. H. Lang puts it, modern 

 phyletic representations more resemble a bundle of sticks than 

 a tree ; the separate lines tend to radiate from a point and not 

 to arise separately from an axis. 



4. That the rate of parallel changes in allied stocks, and the 

 relative rate of these changes of different regions in the same 

 stock, may differ considerably. 



5. That the mode of life is liable to complete changes 

 and reversals : a stock may begin in the water, take to the 

 land, launch out into the air, and then return again to the 

 earth and even end with a life as thoroughly aquatic as that 

 with which it began. Such an animal will retain in its struc- 

 ture features which it has acquired in every stage of its history, 

 most clearly of course those of the later stages, but less and 

 less clearly the results of the adaptations of its earlier 

 ancestors. 



6. That certain insignificant details of structure similar to 

 those which which the Mendelian deals, and on which the 

 systematist founds species, may persist unchanged through 

 considerable changes of the animal's fundamental structure. 



7. That certain features, to all appearance as unimportant 

 as those which are usually regarded as only fit to separate 

 species, not only change in definite directions but seem to 

 retain this trend in all animals whatsoever. The classical case 

 described by C. E. Beecher concerns ornament. Stocks which 

 begin by being smooth, if they develop two sets of ornamental 

 ridges which cut one another will develop tubercles at the 

 intersections, which will subsequently become spines, only to 



