Decembee 17, 1915] 



SCIENCE 



861 



case of nuclear division in any large cell, 

 there is a certain amount of free nuclear 

 division before wall formation begins. In 

 these primitive gymnosperms successive 

 free nuclear divisions continue until nu- 

 merous free nuclei are distributed through- 

 out the egg, and then primary wall forma- 

 tion fills the egg with proembryonic tissue, 

 consisting sometimes of hundreds of cells. 

 The progressive change consists in the 

 earlier and earlier appearance of wall for- 

 mation in the history of the embryo, thus 

 restricting free nuclear division, and limit- 

 ing the extent of proembryonic tissue. 



In the Cycads, for example, permanent 

 proembryonic tissue occurs in every amount, 

 from almost filling the egg (not completely 

 filling it, as in Gingko) to a relatively small 

 amount at one pole of the egg, as in Zamia. 

 When this type of change is followed into 

 the Conifers, the proembryonic tissue is 

 found to be reduced to a few cells, and in 

 some of the Gnetales there is no free nu- 

 clear division, so that the proembryo, in the 

 ordinary sense, has disappeared, a condition 

 which characterizes the angiosperms. The 

 conditions that favor wall formation and 

 inhibit continued free nuclear division are, 

 of course, unknown in a definite way, but 

 that this phenomenon is a response to some 

 progressive change in conditions is evident. 



After recognizing the kind of changes 

 that influence gamete formation, and that 

 perhaps explain the progressive evolution 

 of the archegonium situation, it is of inter- 

 est to discover whether these two series of 

 progressive changes in general proceed pari 

 passu. "Without going into details, it may 

 be said in general that they do. In other 

 words, forms whose arehegonia appear to- 

 ward the maturity of the gametophyte have 

 large proembryos; while those forms that 

 have eliminated arehegonia have also elimi- 

 nated proembryos (that is, in the gymno- 

 sperm sense of free nuclear division as a 



preliminary stage). Whether declining 

 metabolic activity favors wall formation, as 

 contrasted with free nuclear division, as it 

 certainly does gamete formation, I am not 

 prepared to say; but the situation lends 

 itself to experimental answer. 



These two illustrations of progressive 

 evolution suggest that orthogenesis does not 

 differ from other kinds of evolution in being 

 some kind of determinate mechanism that 

 does not respond to a changing environ- 

 ment, but only in that it is a response to 

 some progressive evolution of environment. 

 Of course, we all realize that the word en- 

 vironment covers a tremendous complex of 

 interacting factors, which the ecologist is 

 trying to disentangle. The point made here, 

 however, is not to suggest the factors that 

 have been instrumental in bringing these 

 changes to pass, but to suggest that the fac- 

 tors, whatever they may be, are external, 

 and that the changes are responses. If the 

 change is progressive, the variation in con- 

 ditions is progressive. 



3. The Cotyledons. — No feature of the 

 embryo of gymnosperms has been more dis- 

 cussed than their mixture of dicotyledony 

 and polycotyledony. The discussion has 

 revolved about the conviction that one of 

 these conditions must be primitive and the 

 other derived from it. To some, dicotyled- 

 ony is the primitive condition, because 

 Cycads and Bennettitales are dicotyledon- 

 ous, and they seem most primitive in other 

 features. According to this view polycoty- 

 ledony has been derived from dicotyledony 

 by splitting. To others, polycotyledony is 

 the primitive condition, chiefly because it is 

 characteristic of Abietinete, and in that case 

 dicotyledony has been derived from it by 

 fusion. 



It would be a boon to one or the other of 

 these schools if the embryos of Cycadofili- . 

 cales or Cordaitales should be discovered, 

 and found to be positively either polycoty- 



