216 THE QUANTITATIVE METHOD IN BIOLOGY 



distinguishing the species, recognizing them at first sight. Now 

 we reaUze that the authors of the floras knew the species and 

 even certain subspecies ^ on the whole very exactly. We 

 beheve that the initial difficulties have been a result of our 

 ignorance. Do we realize that they were, above aU, a con- 

 sequence of the hopeless insufficiency of the descriptions ? 



In certain critical genera, such as Rosa, Rubus, Hieracimn, 

 and also Salix, Carex and others, it is true, the difficulties seem 

 to be insuperable. Here the notion of species seems to be really 

 artificial. Let us try to discover the origin of that state of 

 things. 



The prevailing disorder has taken birth in former times, when 

 the disentanglement of the specific forms of the mentioned 

 genera was an insoluble problem. Here the instinct was not a 

 sufficient guide because of the extreme variability. Our pre- 

 decessors had no clear notions about the different sorts of varia- 

 tion. They have described indiscriminately species, subspecies, 

 variants,''' hybrids and bud-variations.^ Since the descriptions 

 were drawn up in the ordinary defective way, it is impossible to 

 find in their writings the exact expression of the observed facts. 

 In more modern times, the same method being still followed, 

 the disorder has been continually augmented by the publication 

 of more memoirs, monographs, and critical revisions. It be- 

 came still more fearful when we began to talk about Jordanian 

 species : the sluice gates were opened more widely than ever 

 before.* The subject has been spoiled. 



A similar disconcerting result would be attained if a number of miner- 

 alogists, ignorant of the MODERN METHODS of investigation of their science, 

 began to write memoirs on the composition of all sorts of sand collected from 

 various parts of the world. 



The investigation of the critical genera of animals and plants should be 

 started afresh. The existing literature should be provisionally thrown aside, 

 the best floras and faunas being used, however, for general orientation and for 

 the discovery of primordia (§ 52). A sufficient number of primordia should 



I Certain forms, described as varieties, are simply variants. (See p. 11.) 



= Numerous species of Carex and Salix are more or less amphibious : it is 

 well known that amphibious species are in general very plastic. 



^ Hybrids and bud-variations may be multiplied indefinitely by asexual 

 propagation. (See § 20, p. 16.) The latter is widely spread in the genera under 

 consideration. 



FR. CRliPIN, who devoted more than thirty years of his life to the study 

 of the genus Rosa, has called attention to the fact that many so-caUed species 

 of that genus have been based upon one specimen, or upon one bush [buisson) 

 consisting of specimens which were all produced by vegetative propagation of 

 a single one. Such species are represented in numerous herbaria by branches, 

 which were distributed by exchange clubs or exsiccata. CR^PIN called this 

 method buissonomanie. 



A similar method has been probably appUed on Rubus, Salix, Carex, etc. 



' Investigation of the Jordanian species ought to be carried out by means 

 of the quantitative method initiated by JOHANSEN for the pure lines in his 

 experiments with beans. 



