752 



SCIENCE. 



[N. S. Vol. VII. No. 178. 



of species, I cannot follow him at all. Size and 

 weight — the traits that can be measured — are 

 especially dependent on the environment and 

 variable within the same species. Varieties of 

 dogs may not intergrade at all in size and weight, 

 or in the relative dimensions of the skeleton, 

 but this does not lead us to call them separate 

 species. The cephalic index is one of the most 

 important differentials in man, but the fact that 

 it may not intergrade does not turn races into 

 species. The conditions are far more complex 

 than Mr. Davenport assumes them to be. A 

 certain quantitative amount of intergrading 

 may mean entirely different things under dif- 

 ferent circumstances, and the various differen- 

 tials of a species may intergrade to very dif- 

 ferent degrees. It does not follow that the chief 

 ■differential is that quantitative characteristic 

 intergrading the least. It may be the teeth or 

 the reproductive system or whatever serves 

 most conveniently as a basis of classification. 



My excuse for writing on the definition of 

 species is that I hold it to be a psychological 

 problem. In pre-evolutionary days the natural- 

 ist undertook to discover species that had been 

 created ; now it is he who creates the species.* 

 The problem is analogous to deciding how many 

 •colors there are in the spectrum; it may be held 

 that there are three, or four, or seven, or two- 

 hundred. There are, indeed, various criteria 

 that may be used in the separation of species, of 

 which the most important seem to be : (I) the 

 phylogenetio history when known ; (2) heredi- 

 tary stability and variability ; (3) the tendency 

 to cross and the fertility of crosses, and (4) in- 

 tergradation. The last named factor is not only 

 quantitative, as in the cases given by Mr. Daven- 

 port, but also qualitative, and here the natural- 

 ist must try to use as his unit what the psychol- 

 ogist calls the 'just observable difference.' The 

 degree of distinctness that shall constitute a 



*1 fear that I am here sailing under Dr. Merriam's 

 heavy guns. He has -written : "The function of the 

 naturalist is neither to create nor destroy species, but 

 to recognize, describe and learn about those which 

 nature has established." (Science, N. S., V., 124.) 

 Innumerable coyotes, differing more or less, live or 

 have lived, and Dr. Merriam, not nature, has estab- 

 lished eleven species. Some other naturalist has 

 created the coyotes. 



species must, like the meaning of every word, 

 depend on the best usage. As the usage of the 

 best writers is compiled and given currency by 

 dictionaries, so the usage of naturalists is com- 

 piled and given currency in a work such as Das 

 Thierreich. The criterion given prominence by 

 Messrs. Davenport and Blankinship should be 

 carefully studied, but it is only one of many 

 factors, and these must be distinguished and 

 adjusted by the powers of observation and 

 judgment of the naturalist. The definition of 

 species is, as I have said, a psychological prob- 

 lem. 



J. McKeen Cattell. 



SCIENTIFIC LITERATURE. 

 Contribution towards a Monograph of the Lahoul- 

 heniacese. By Roland Thastbe. Memoirs 

 of the American Academy of Arts and Sci- 

 ence. 1896. Vol. XII., No. 3. Pp. 189- 

 429. 26 plates. 



This is the second important memoir by Dr. 

 Thaxter on Entomogenous fungi, the first being 

 a monograph of the Entomophthorese. The 

 very large number of these plants which are 

 being brought to light by the keen observation 

 and untiring industry of the author of this 

 memoir is a surprise to any one acquainted with 

 the literature of the subject. 



As Dr. Thaxter states in the introduction, his 

 study of Entomogenous fungi was begun with 

 the intention of embodying in a single mono- 

 graph all species truly parasitic on insects. But 

 the number of species of the Entomoythoreje 

 were sufficient for a monograph of considerable 

 proportions, and now the hitherto insignificant 

 family of Laboulbeniacese has, under his ind- 

 fatigable researches, grown to an order of for- 

 midable proportions, while several other groups 

 of insect fungi remain yet to be investigated. 

 While a few of the members of the genus La- 

 houlbenia have been known for nearly one-half a 

 century, our knowledge of the development, sex- 

 uality and formation of the spores has remained 

 very imperfect. This, together with the diffi- 

 culty of defining the position of the family in 

 relation to other thallophytes, has probably had 

 much to do with the almost universal absence 

 of treatment of these forms from text-books of 

 funari. 



