1885.] NATURAL SCIENCES OF PHILADELPHIA. 309 



groups he considers such difference to be at least of generic im- 

 portance, and what is more singlar, he even constructed thereon 

 a whole family. His Melocrinidae differ from his Actinocrinidae 

 mainly in having all five inter radial spaces arranged almost uni- 

 formly, and they generally have Pour basals. That Miller did not 

 make the number of basals the distinctive character, is very evi- 

 dent, or he would have arranged his Xenocrinus penicillus and 

 " Compsocrinus" Harrisi among the Melocrinidae. On the con- 

 trary, he placed these species under distinct genera; while he re- 

 ferred Glyptocrinus decadactylus and Reteocrinus Richardsoni to 

 the same genus, although these two differ in exactly the same 

 way as the two former species. 



One is curious to know upon what ground Miller based his 

 Glyptocrinidae. Not upon the underbasals, nor upon the relative 

 number of basals ; neither upon the riclges along the radials, for 

 these are absent in Cupulocrinus and Lampterocrinus, and cer- 

 tainly not upon the ornamentation, which he asserts does not 

 hold good even among those genera. They are united by no 

 single character, and since it has been clearly proved that Glyp- 

 tocrinus has no underbasals, this genus no longer falls within the 

 Rhodocrinidae, which were fundamentally based upon the pres- 

 ence of those plates, and must be referred to the Actinocrinidae. 

 That Glyptocrinus was in many respects closely allied to the Ac- 

 tinocrinidae, subdivision Melocrinites, has been shown already 

 in Part II, and several species were at first described under Glyp- 

 tocrinus, which we have since referred to Mariacrinus. Among 

 these is Gl. Harrisi, for which Miller lately proposed the genus 

 Compsocrinus. The generic definition of Gompsocrinus is partly 

 based upon inaccurate observation, for the interradials of all five 

 sides rest upon the edges of the first radials, and not one of them 

 upon a basal, as figured by Miller in his diagram pi. 11, fig. 4, a. 



It has been stated in Part II, p. 185, that the interradials of all 

 known Actinocrinidae, except sometimes those of the azygous 

 side, rest upon the first radials. and this is the case in Glyptocri- 

 nus and "Compsocrinus." We find an apparent exception to this 

 rule, if we make the absence of underbasals the controlling family 

 character, in t lie genus Xenocrinus, and perhaps in Glyptocrinus 

 Richardsoni and Gl. Patter son i, in which underbasals have not 

 been observed. The two latter species agree in all essential par- 

 ticulars with Xenocrinus, in which we include not only Miller's 



