64 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



species Avill be split up. Long ago Sladen* pointed out the possible 

 error in too comprehensive species, such as poly acanthus. He said: 



Without calling in question tlae accuracy of M. Perrier's determination, tlie 

 occurrence of sucli instances as this of a form presenting strongly marked 

 variations at different stations within the area of its distribution, urges upon 

 naturalists the necessity of exercising extreme caution against being led away 

 by a tendency to group too comprehensively the forms which may be included 

 within a large and widely distributed genus; for, however seriously the mul- 

 tiplication of frivolous " species " may embarrass a classification, the whole- 

 sale grouping, or, in other words, the unbounded extension of the limits of 

 specific character, is productive of more injurious results, in that it curtails 

 the precision of definition, and. whilst ignoring environment as a factor, 

 divests nomenclature of one of its highest and most important qualities. 



From the fact that forms are separated by much smaller and less striking 

 differences in an extensive genus, than in one of more limited scope, " species " 

 in the larger group have often not such clearly marked or conspicuous charac- 

 ters as those which are presented by " varieties " m a less comprehensive genus. 

 It follows that the judgment should be very cautiously exercised when tempted 

 to embrace within a single species all the strongly marked distributional ex- 

 tremes of any widely spread type, however closely their connection may seem 

 to be preserved through intermediate forms ; for in many cases these grada- 

 tions are nothing more or less than the links which indicate to us the develop- 

 ment of •' species," and are, in short, the stages with which generally we are 

 unacquainted, owing either to the imperfection of knowledge, or more fre- 

 quently by reason of their destruction through the hostility of unfavorable 

 conditions. 



Without a large series of specimens from numerous localities it 

 will be difficult to arrive at a solution of AHropecten polyacanthus, 

 for it is likely that the rather conspicuous individual and habitat 

 variations somewhat mask small and constant differences due to 

 wide geographic separation. 



Type-locality. — Red Sea. 



Distribution. — The Red Sea to Zanzibar and Mozambique, the Sey- 

 chelles, Ceylon, Mergui, Andaman Islands, China, Japan, Philip- 

 pines, Port Jackson, Australia, Admiralty Islands, Aru Islands, 

 Fiji Islands, Hawaiian Islands. 



Specimens examined. — Six specimens from the following localities : 



Catbalogan, Samar, shore; 1 specimen. 



Usada Island, vicinity of Jolo, shore ; 1 specimen. 



Station 5159, Tawi Tawi Group, Sulu Archipelago, 10 fathoms, 

 coarse sand, shells; 1 specimen. 



Station 5165. Same locality, 9 fathoms, coral; 2 specimens. 



Station 5174. Vicinity of Jolo, 20 fathoms, coarse sand; 1 speci- 

 men. 



iJourn. Linn. Soc. Zool., voL 14, 1879, p. 429. 



