THE SPECIES CONCEPTION. XXV 



the same colony or less frequently in separate colonies. The 

 direction of coil is hereditary as a general rule. I have never 

 found a sinistral embryo in the uterus of a dextral mother. 

 Mr. Thaanum, whose experience has been very wide, states 

 that he has never taken both dextral and sinistral embryos 

 from one mother, in the few species of Partulina which are 

 indifferent in coil. 



Frequently in adjacent colonies of the same race, sinistral 

 shells may predominate in one, dextral in another. In other 

 species, the prevalent direction of coil differs in different 

 parts of its area. See Partulina virgulata, p. 358, Acha- 

 tinella vulpina, and many other cases noted in the systematic 

 part of this work. It seems likely that, as Mr. Gulick thought, 

 dextral and sinistral snails find some mechanical difficulty in 

 mating; but experimental evidence is needed. 



THE SPECIES CONCEPTION IN ACHATINELLA. 



The conception of "species" applied to Achatinella is in- 

 volved in more than the usual difficulty. We have to do with 

 evolution products of several grades, possibly of more than 

 one kind. 



Grade I. Races having certain seemingly permanent as- 

 sociations of characters (though usually with other variable 

 characters), and co-existing with allied races without form- 

 ing hybrid colonies. 



Grade II. Geographic races (i. e. with a definite and con- 

 sistent distribution ) , having moderately coherent associations 

 of characters, but blending with other like races through 

 hybrid or undifferentiated colonies where the geographic 

 ranges meet. 



Grade III. Forms characterized by the possession of one 

 or more special characters (usually of color), but which occur 

 for the most part in hybrid colonies with other diverse forms ; 

 the characters seeming to be freely interchangeable, and oc- 

 curring in different combinations, as in Mendelian hybrids. 



The races included under (I) are species in the usual sense. 

 It may be supposed that the specific isolation is physiologic. 

 Thus, Achatinella bellula and A. vulpina are related species 



