296 



SCIENCE. 



[N. S. Vol. XXIII. No. 582. 



bles that found in the vegetative stems of 

 some of the very ancient Cycadofilices. 

 Similarly the present speaker has observed 

 that the structure of the woody axis of the 

 cone in living species of Pinus differs 

 strikingly from that found in the vege- 

 tative stem, and strongly resembles that 

 found in Cretaceous Pityoxyla. This im- 

 portant principle of the persistence of an- 

 cestral features in reproductive axes is 

 particularly significant, because it offers 

 an independent support for the time-hon- 

 ored practise of the systematic botanist, 

 who attaches great importance to the floral 

 structures and their arrangements, in tra- 

 cing lines of affinity in the flowering plants. 



Another important new phylogenetic 

 principle, which has recently emerged, and 

 which is likewise the special property of 

 the botanical morphologist, is that ancestral 

 characters are extremely likely to persist 

 as structural features of the leaf. For ex- 

 ample, the leaf-bundles of the cycadean 

 gymnosperms are the exact counterpart of 

 the stem-bundles of some of the extinct and 

 probably ancestral Pteridosperms. This 

 principle might also be illustrated by many 

 examples if time permitted. 



We have thus three important phylo- 

 genetic laws resulting from our more com- 

 plete knowledge of the older vegetation of 

 the earth. These principles or laws having 

 been elucidated by the comparison of living 

 with fossil forms may now fruitfully be 

 extended as general working rules to the 

 comparison of living groups with one an- 

 other. The importance of these principles 

 can scarcely be overestimated; for they 

 enable us at once to put the sporophytic 

 generation in the foreground as the basis 

 of phylogenetic study. This is particularly 

 fortunate, because it is precisely the sporo- 

 phyte which allows fruitful comparison 

 with extinct forms, since the gametophyte 

 does not ordinarily become fossilized. More- 

 over, since the time of Hofmeister, the 



gametophytic generation has performed 

 such an important role in morphological 

 investigations that in recent years, in spite 

 of the important discoveries of chalazog- 

 amy, the antherozoids of the cycads and 

 Ginkgo, and double fertilization in the 

 angiosperms, it has begun to show signs of 

 exhaustion. The next half century, with- 

 out neglecting the gametophyte or the 

 earlier stages of the sporophyte, will doubt- 

 less give a great deal more attention to the 

 later development and mature structure of 

 the sporophyte, which being the predom- 

 inating and unreduced generation, in the 

 vascular plants, will yield, as our knowledge 

 of the ancient forms becomes more com- 

 plete, the most important morphological 

 and phylogenetic results. 



A further fruitful field for morpholog- 

 ical exploitation is that of experimental 

 morphology. This field, although much 

 discussed and canvassed at the present 

 moment, is as yet practically untouched 

 from the phylogenetic side. It is diffi- 

 cult to see how it can be successfully culti- 

 vated by those who have not a reasonably 

 complete knowledge of what may be called 

 the normal anatomy of living and fossil 

 plants. We appear as yet to be no nearer 

 to the possibility of experimentally orig- 

 inating new species. Indeed, if we ever 

 succeed in penetrating the veil with M'hich 

 nature conceals this part of her workings, 

 the hypothesis that acquired charaetera can 

 not be transmitted will have to be given up. 

 For if by experiment we are able to bring 

 it about that species acquire and transmit 

 new characters and thus become new spe- 

 cies, the doctrine of the non-transmission of 

 acquired characters will become ipso facto 

 obsolete. 



In conclusion, it may be said that there 

 appears to be no immediate prospect that 

 the practise of making genealogical or phy- 

 logenetic trees will have to be abandoned. 

 In constructing these trees, however, we 



