PROBLEMS OF PLANT MORPHOLOGY 89 



So, as the matter now stands, we cannot deceive ourselves on this 

 point, that the constructions of the old morphology, although con- 

 fined almost entirely to vestigial series, nevertheless stood on firmer 

 ground than the modern speculations on the question of primitive 

 forms. Starting with a completely endowed form, we can follow 

 the reduction of form through intergradations, and, by reference 

 to vestigial organs, often with convincing certainty. But by what 

 means shall we judge a rudimentary organ? Is it more than a gratui- 

 tous assumption, when, as recently was the case, a certain botanist 

 declares the lodicules of grasses to be not a perigone, but a rudiment 

 of a perigone? * Whereby may one recognize a rudiment, i. e., the 

 attempt to form something new, an attempt which, however, has 

 remained nothing more? In what way may we distinguish such a 

 rudiment from a vestigial organ? And, finally, after one has broken 

 faith with the old vestigial series, is it not still more of the stamp of 

 formal morphology if he contents himself in arranging forms in series 

 and then comes to a standstill when he tries to decide at which end 

 stand the primitive and at which end the derived forms? At any 

 rate, such a limitation brings out the better the true condition of our 

 knowledge, for such an arrangement of forms in a series is about 

 the best service that formal morphology can do. This service is, to 

 be sure, no small one, for it enhances broad critical comparison, 

 and is, therefore, the result of hard work. But the desire to give 

 this arrangement in series a genetic bearing has oftentimes led us 

 to untenable propositions and explanations. Just as we have little 

 ground for assigning the Cupuliferce to a primitive position, so have 

 we as little evidence for regarding the Casuarince also in the same 

 light. The latter have been placed by a recent systematist at the 

 apex of his system, because there has been an inclination to find in 

 them a sort of "missing link" between angiosperms and gymno- 

 sperms. 2 I may, perhaps, mention that I had regarded such a view 



1 Engler, Die systematische Anordnung der monocotyledonen Angiospermen. (Ab- 

 handl. der K. Preuss. Akad. d. Wissensch. zu Berlin, p. 22.) The reasons advanced 

 by Engler for the view that the primitive flower of the Graminece was naked are 

 quite beside the mark, as, e. g., when he expresses the opinion that wind pollination 

 indicates that the types of the Graminece and Cyperacece are very old. Is, then, 

 the Plantago type very old, or Thalictrum "older" than the other Ranunculece? 

 We know that wind pollination has appeared in widely different groups of plants, 

 evidently in part as a reduction of flowers which were not wind-pollinated. A cor- 

 relation between the flowers and the glumae, or palese, puts Engler on the defense. 

 We see, however, very plainly in many grasses that leaf-organs which have become 

 superfluous as a result of their position, have become reduced. A similar process 

 may have taken place in the perigonal leaves, and the behavior of Streptochceta 

 strongly indicates this. (Cf. on this point Goebel, Ein Beitrag zur Morphologic der 

 Graser, Flora, 81. Bd. Erganzungsbd. z. Jahrg. 1895.) The seed and fruit characters 

 of the grasses also are anything but primitive. 



2 Engler (Syllabus, Fourth Edition, 1904) has recently yet again placed the 

 Casuarince at the point of divergence of the Dicotyledonce, although the conten- 

 tion that in their macrospore before fertilization " a rudimentary prothallium, 

 consisting of twenty or more nuclei, arises" cannot be maintained by the earlier 

 researches of Treub, as I have pointed out (Organographie, p. 894). Frye (The 



