340 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



latter not subdivisions into which the Avhole of the group is split. 

 In classifying individual organisms it will thus often happen that 

 the group (species) is the only category we can use, but in other 

 cases the individuals: conform to well-marked types and may there- 

 fore be referred to varieties. This is the practice which F. E. 

 Schulze years ago marked out as the proper one (1879, pp. 11-12) 

 when dealing with exuberant variability, and, I believe, it is under- 

 stood by many to be the correct procedure. Nevertheless in S3 r s- 

 tematic manuals of to-day we frequently find a widely ranging and 

 variable species represented as separable into subgroups, usually 

 designated subspecies, the implication being not that one is a rela- 

 tively heterogeneous remainder but that all are equivalent groups 

 stamped with about the same degree of homogeneity. 



With regard to a fundamental difference in meaning between 

 terms, such as subspecies and variety, for intraspecific groups, all 

 systematists must confess that our knowledge of variation in sponges 

 (and organisms in general) is yet too fragmentary for us to apply, 

 widely and consistently, in the work of classification the ideally 

 different categories for which such terms are sometimes meant to 

 stand. The increased use in recent years of " subspecies '" has been 

 marked, especially for geographical races, and this has perhaps 

 tended to crystallize the erroneous idea that a species is a congeries 

 of equivalent smaller groups, into some one of which any particular 

 individual must go. 



Nor are we as yet able to use, except here and there, in systematics 

 the various kinds of intraspecific groups with which experimental 

 genetics begins to make us acquainted. As to the genetic values and 

 correlations of the differential characters that stamp varietal types 

 we are largely in the dark. This is true, for instance, of the dif- 

 ferences in cortical anatomy which mark off D. inr/alli, sens str., from 

 D. seychellensis, as it is true of the absence of a particular microsclere 

 (as in D. japonica). Topsent (1918) tends to disregard the fea- 

 tures of cortical anatomy. But, as far as I know, there are no 

 recorded data which indicate that they are less constant than spicular 

 features. 



Genus TUBERELLA Keller (1880). 



Tubcrelhi Keller, 1SS0. p. 270. 

 Without a fibrous cortex and without microscleres. Chief megas- 

 cleres are large fusiform styles arranged in distinct spiculo-fibers 

 that course toward the surface, or which are so abundant that an ar- 

 rangement in fibers is scarcely perceptible. Small slender styles 

 radiateh T arranged occur in the ectosome. 



