September 27, 1901.] 



SCIENCE. 



473 



disclose new and as yet unsuspected rela- 

 tions to forces controlling distribution in 

 general, but they would give us our first 

 accurate landmarks for the observation of 

 plant migration, thus greatly facilitating a 

 study of progressive changes in our flora. 



While the accurate determination of plant 

 boundaries has thus great interest it may 

 be remarked that research in this field, as in 

 others, to be successful, must be conducted 

 with care. Keports of occurrence, especially 

 extra-limital stations, should be taken with 

 much caution. In this as in many other 

 matters of science it is impossible to make 

 too sharp a distinction between facts actu- 

 ally observed and those taken on hearsay. 

 In mapping a plant the recorder will do 

 well to indicate this difference. If, for in- 

 stance, he shows by an umbra the range 

 which rests upon specimens personally ex- 

 amined, let him record unverified reports 

 only by a penumbra. Furthermore, any 

 work of this kind to be of permanent value 

 must rest, at least in great part, upon speci- 

 mens which are carefully preserved, for 

 segregation is progressing rapidly and no 

 one can foresee its subjects. A plant of 

 supposedly uniform character may at any 

 time prove, upon more critical observation, 

 to be two or more distinguishable species. 

 In such a case it is easy to see that any 

 previous studj^ or records of the composite 

 plant must lose nearly all their value unless 

 specimens have been preserved so that are- 

 examination will show to which of the seg- 

 regates the records applied. Similarly, the 

 disappearance of a plant from a given region 

 may lead to a justifiable scepticism as to the 

 accuracy of the records relating to its occur- 

 rence in that place. In this case, practi- 

 cally the only valid proof is a preserved 

 specimen accompanied by the original data 

 of collection. 



In interpreting and recording this matter 

 of plant boundaries opinions will doubtless 

 dififer as to what may be called continuous 



range and what is to be regarded as an extra- 

 limital station, or, so to speak, an island in 

 a sea of non-occurrence. This is, of course, 

 all a matter of degree, since in reality no 

 plant has a continuous range, for it is rep- 

 reseated by more or less isolated individ- 

 uals. Yet this ofiers no serious obstacle. 

 The meteorologist maps the analogous 

 course and limits of a rain-storm composed 

 of separate drops, and the biologist has long 

 recognized the practical continuity of plant 

 and animal ranges which, in a generalized 

 form, are the basis of his so-called ' life- 

 zones.' 



Turning now to quite a different field 

 which seems to offer great possibilities, I 

 would call attention to recent researches in 

 plant ontogeny : the investigation of em- 

 bryonic development, the comparative 

 study of seedlings, and such observations 

 as have been recently made by Professor 

 R. T. Jackson upon the reappearance of 

 juvenile and ancestral traits in offsets and 

 runners. Systematic zoologists have long 

 made use of ontogeny in determining group 

 aflBnities, but botanical taxonomists have 

 been much less successful in drawing from 

 the early stages of plants like inferences. 

 There are several reasons for this. In the 

 first place, there can be no doubt that 

 plants in their early development do not 

 exhibit such a continuous and complete 

 series of philogenetic stages as many 

 animals do. In plants some stages seem 

 to have dropped out by a sort of morpho- 

 logical and physiological elision or ellipsis. 

 Again, while the classification of animals 

 rests upon general morphology often well 

 suggested even in very early stages of de- 

 velopment, the classification of plants is 

 based chiefly upon the mode of reproduc- 

 tion — that is to say, upon a series of struc- 

 tures produced so late in the life of the 

 individual that no suggestion of their char- 

 acter is afforded by embryo or seedling. 



But, after all, there can be no doubt that 



