782 



SCIENCE 



[N. S. Vol. XLII. No. 1092 



parallel series in this group as in the 

 Algse. The Pteridophyta seem a better 

 case, for we have direct evidence from 

 fossil plants as well as the comparison of 

 living forms to assist us. Though paleo- 

 botany has added the Sphenophyllales to 

 the existing groups of vascular crypto- 

 gams and has greatly enlarged our concep- 

 tions of the others, there is no proof of 

 how the great groups are related to one 

 another. As in the Bryophyta, they may 

 represent several completely independent 

 parallel lines. There is no evidence as to 

 what sort of plants the Pteridophyta were 

 derived from, and in particular none that 

 relates them to any group of Bryophyta 

 or Algfe. I do not want to labor the argu- 

 ment, but much the same can be said of 

 the seed-plants, though there is consider- 

 able evidence and fairly general agree- 

 ment as to some Gymnosperms having 

 come from ancient Filicales. The progress 

 of phyletic work has thus brought into 

 relief the limitations of the possible results 

 and the inherent difficulties. As pointed 

 out by Professor Bower, we can hope for 

 detailed and definite results only in par- 

 ticularly favorable cases, like that of the 

 Filicales. 



The change of attitude shown in recent 

 phyletic work towards "parallel develop- 

 ments in phyla which are believed to have 

 been of distinct origin" is even more sig- 

 nificant. Prof. Bower spoke of the preva- 

 lence of this as an "obstacle to success," 

 and so it is if our aim is purely phyletic. 

 In another way the demonstration of par- 

 allel developments constitutes a positive 

 result of great value. Thus Professor 

 Bower's own work has led to the recogni- 

 tion of a number of series leading from 

 the lower to the higher Filicales. By inde- 

 pendent but parallel evolutionary paths, 

 from diverse starting-points in the more 

 ancient ferns, such similarity has been 



reached that systematists have placed the 

 plants of distinct origin in the same genus. 

 In these progressions a number of charac- 

 ters run more or less clearly parallel, so 

 that the final result appears to be due "to 

 a phyletic drift that may have affected 

 similarly a plurality of lines of descent." 

 This conclusion, based on detailed inves- 

 tigation, appears to me to be of far-reach- 

 ing importance. If a "phyletic drift" in 

 the ferns has resulted in the independent 

 and parallel origin of such characters as 

 dictyostely, the mixed sorus, and the very 

 definite type of sporangium with a ver- 

 tical annulus and transverse dehiscence, 

 the case for parallel developments in other 

 groups is greatly strengthened. The in- 

 terest shifts to the causes underlying such 

 progressive changes as appear in parallel 

 developments, and the problem becomes 

 one of causal morphology rather than 

 purely historical. 



The study of parallel developments 

 would, indeed, seem likely to throw more 

 light on the morphology of plants than the 

 changes traced in a pure phyletic line, for 

 it leads us to seek for common causes, 

 whether internal or external. We cease to 

 be limited in our comparisons by actual 

 relationship, or forbidden to elucidate the 

 organization in one group by that which 

 has arisen independently in another. 

 Similarly the prohibition against compar- 

 ing the one generation in the life-cycle 

 with the other falls to the ground, quite 

 apart from any question of whether the 

 alternation is homologous or antithetic. 

 The methods of advance and the causal 

 factors concerned become the important 

 things, and if, for example, light is thrown 

 on the organization of the fern-plant by 

 comparison with the gametophyte of the 

 moss, so much the better. This, however, 

 is frankly to abandon phylogeny as "the 

 only real basis of morphological study," 



