August 11, 1905.] 



SCIENCE. 



185 



established by the discovery of a more com- 

 plete series of transition stages. If, on the 

 other hand, the forms to which so much sig- 

 nificance has been attached are merely adapta- 

 tions to peculiar life conditions, assumed by 

 a plant at an exceedingly plastic stage of its 

 development, the problem of the relationship 

 of the two great groups seems to be farther 

 than ever from a satisfactory solution. 



The value of seedling characters for the 

 tracing of phylogenetic development in fam- 

 ilies and lower systematic groups and an un- 

 derstanding of past changes both in the race 

 and in its environmental conditions, has been 

 emphasized by several writers. The most 

 satisfactory results in this field are those ob- 

 tained by Ganong and Cockyane, who have 

 both worked principally with xerophytic 

 forms which offer especially promising ma- 

 terial for such researches, though here ex- 

 treme caution must be exercised in deciding 

 whether distinctive characters of the adult 

 are not merely an expression of the direct in- 

 fluence of the environment upon a plastic 

 organism, while the juvenile stages, owing to 

 the different conditions of growth, are not 

 subjected to these influences, and consequently 

 all the potentialities of their primordia may 

 be realized instead of intercepted and diverted 

 or modified by powerful environmental influ- 

 ences. It scarcely need be suggested that 

 here experimental morphology and ecology 

 have before them material for a series of very 

 interesting monographs. 



A field in which results of especial interest 

 may be expected is the comparative investiga- 

 tion of the later developmental stages within 

 the same systematic group. For families 

 Ganong has presented a masterly treatment of 

 the Cactacese and Willis has accumulated 

 some data for, and suggested the importance 

 of, such work in the Podostemacese, while 

 Miss Sargent's work on the Liliacese, when it 

 appears, will doubtless represent the most ex- 

 tensive study of the kind ever undertaken. 

 In other families there is a large store of de- 

 tails waiting for supplementary researches 

 and correlation. The results so far obtained 

 show beyond question the interest, ecological. 



morphological and phylogenetic, attached to 

 the study of these groups. While I would not 

 discourage the investigations of these larger 

 groups, it must be admitted that in the present 

 state of our knowledge there are many uncer- 

 tainties connected with the generalizations 

 concerning their phylogeny, and for the pres- 

 ent especial importance should be attached to 

 the investigation of the minor groups, par- 

 ticularly the genera. We have little right to 

 assume the monophyletic origin of the most 

 of the families, while with the genus this is 

 more justifiable, though even here the greatest 

 caution must be used. The embryonic stages, 

 especially the later ones, should be of the 

 greatest value in just this connection, for 

 theoretically they ought to furnish us with an 

 indication of form prevailing prior to the 

 assumption of the more specialized adaptive 

 characters. 



But before we may draw conclusions as to 

 the phylogeny of a group or conclude whether 

 monophyletic or polyphyletic in its origin, 

 from characters offered by the later embryonic 

 stages of its members, we must first under- 

 stand thoroughly the seedling and its reaction 

 to all the factors influencing it. This is a 

 field not for comparative work alone, but for 

 physiological, ecological and experimental 

 morphological investigations as well. Com- 

 parative stu.dies are of the highest importance, 

 but in this case they must be carried out upon 

 material of systematically homogeneous na- 

 ture. Until the appearance of Professor de 

 Vries's epoch-making work no one has had 

 available for study a series of forms unques- 

 tionably descended from the same ancestor. 

 It is highly desirable that some one take up 

 such types as the new elementary species of 

 Professor de Vries for the purpose of ascer- 

 taining in how far there is a relation of an- 

 cestral characters and to what extent there is 

 in the seedling a working back into the em- 

 bryo of the characters of the adult. While 

 the new species of CEnoihera described by 

 Professor de Vries are the simplest systematic 

 units, there are some very suggestive points 

 to be found in his descriptions. The relative 

 stages of development at which the differential 



