230 



THE AMERICAN NATURALIST 



[Vol. XLI 



predominates in a region, in other regions several groups are 

 nearly equally represented. Within each group, divisions can 

 be made; but in the case of these divisions geographic sepa- 

 ration does not obtain, since species of all the divisions of a group 

 are Hkely to occur in any part of the general territory proper to 

 the whole group. Regarding the ultimate units, or species, those 

 which are most closely allied are likely to be found promiscuously 

 associated in the same district and without the semblance of isola- 

 tion. For example species of the Pruinosse or of the Intricatse 

 with 10, or with 20 stamens, or with rose-colored, or with yellow 

 anthers are found growing within a few feet of one another, and 

 may cover common districts of several hundred square miles. 

 In these cases, while it is the number of stamens or color of anther 

 which first attracts attention, other specific characters exist which 

 adequately distinguish the species. As an example of promiscuous 

 association, the vicinity of Albany may be pointed out, where 

 the five species of Intricatse heretofore found in New York state 

 grow in a small area. In Ontario we find twenty-five species of 

 Tomentosse, many of them growing very close together. In the 

 distinctly southern group Microcarpse we find the two species, 

 C. apiifolia and C. spathulata, growing over the same areas, while 

 the third and more distantly related species, C. cordata has a some- 

 what more northern range. In general, the reverse of Jordan's 

 law would more nearly represent the distribution of American 

 species of Crataegus. 



Coming now to Orchidacea?, I may say that I adopted the line 

 of examination suggested by the form of Jordan's law; that is, 

 I looked for pairs of kinds. I say kinds instead of species in- 

 tentionally. The main problem should not be confused by the 

 difficulty of agreeing upon a definition of species. What the evo- 

 lutionist has to account for is not the definitions of systematists, 

 but the multipHcity of hereditary types; he has to explain the 

 antithesis between the uniformity which heredity seems at first 

 to promise, and the diversity which actually prevails among 

 organic things. A definition of species is demanded in taxonomy, 

 but is somewhat less necessary in studies like the present. We 

 do not require that the foims be related in some particular taxo- 

 nomic sense; but only that they have different hereditary charac- 



