July 14, 1905.] 



SCIENCE. 



37 



are we now concerned — beyond the mere 

 descriptive consideration of development is 

 here of significance, and brings the possi- 

 bility of further progress. And small in- 

 deed, I may add, appears to be such ad- 

 vance to those who, from the beginnings of 

 phylogenetic morphology have, like Sisy- 

 phus, sustained their courage to roll again 

 and again up the mountain the rock of 

 phylogeny as often as it has rolled down. 

 It may now be attempted to examine 

 somewhat more closely in certain particular 

 examples the relation between phylogenetic 

 and causal morphology. One of the 

 changes which phylogenetic morphology 

 has brought with it is that it seeks to ascer- 

 tain which form is 'primitive' and which 

 derived. Idealistic morphology has borne 

 in upon us no conviction on this ques- 

 tion, since it derives all forms from a type 

 which is present only as a conception. 

 But phylogenetic morphology must, on the 

 one hand, always reckon with the possi- 

 bility of polyphyletic development, and, on 

 the other hand, it can operate not only with 

 reversionary structures, as did the ideal- 

 istic morphology, but must be far more 

 concerned in determining which forms 

 within the series which it proposes stand 

 nearest the common point of derivation. 

 It seeks then with diligence after 'primi- 

 tive' forms. But in this search we meet 

 with great difficulties. In the first place, 

 we are inclined to regard those forms as 

 primitive which have simple form-rela- 

 tions, and unmarked division of labor. 

 But such forms may also have arisen by 

 reversion, and if one looks over botanical 

 literature, lie sees, at least so far as the 

 relationships between the larger groups are 

 concerned, there exists no agreement as 

 to which forms are to be regarded as primi- 

 tive and which derived : often opinion on 

 this point changes with the fashion. Thus 

 the thallose liverworts have up till now 

 been regarded as more primitive than the 



foliose, because the vegetative body of the 

 former is much more simple in construc- 

 tion than that of the latter, and between 

 them there are found gentle gradations. 

 Recently, however, the attempt has been 

 made to derive the thallose from the foli- 

 ose forms. This is not the place to examine 

 the evidence for or against such derivation. 

 How vacillating is the point of view from 

 which it is judged what form is primitive 

 is shown by the various positions which 

 have from time to time been given to the 

 apetalous dicotyledons. 



The old morphology regarded these as 

 reduced forms because their flowers are 

 less fully differentiated than those of most 

 of the other dicotyledons. Eichler has, 

 however, already shown that there is no 

 ground for maintaining that the corolla in 

 the 'luliflorse' and ' Centrospermas ' has 

 suffered reduction; and on this point we 

 can only agree with him. But must they, 

 because the perianth shows simpler form 

 relations and also because the number rela- 

 tions within the flower are not always con- 

 stant, be therefore primitive ? Even if we 

 admit that these groups have a great geo- 

 logical age, it is not proved that they stand 

 as regards their total organization on a 

 lower plane of development ; old and primi- 

 tive forms are the same only when it can be 

 shown that the former stand nearer to the 

 stem forms of the angiosperms than other 

 forms. If this is not capable of proof, 

 then the old forms may just as well be the 

 end terms of long developmental series as 

 others, only that the differentiation of 

 organs has not taken place to the same 

 degree as in the others. Now, we do not 

 Ivuow the stem forms if the angiosperms, 

 and they may never, perhaps, be known. 

 But even if we content ourselves by recon- 

 structing them on the basis of comparative 

 study, I can find no reason, e. g., to regard 

 the Cupuliferffi as primitive forms, while 

 I can find many reasons for not doing so. 



