1 82 SMITHSONIAN INSTITUTION 



most of the rays a normal sized axillary, and beneath it a short, thin plate which may be con- 

 sidered an immature or suppressed primibrach (PI. XI, fig. na). 



In the English form, otherwise very similar to some forms of the type species, and 

 another closely related species from Tennessee, the number of primibrachs is constant at one 

 in at least 15 specimens. But the Tennessee collections have yielded a thoroughly distinct 

 species having three and four. In another Gotland specimen which I consider an abnormal 

 P. nodulosus there is another irregularity similar to one of those noted in Lecanocrinus ; the 

 radianal appears in primitive position under the right posterior ray, and the anal x has been 

 pushed out of the calyx, being represented by two or three small plates above the level of the 

 radials (PL XI, fig. 166). 



Pycnosaccus is notable as being one of the genera which lack the characteristic enlarge- 

 ment of the upper part of the column so common in this group ; the stem is constructed very 

 much like that of Lecanocrinus. Thanks to Mr. Liljevall's skillful preparation, we are able to 

 add this to the list of genera in which the structure of the tegmen is known. His fine draw- 

 ings from a specimen of P. nodulosus show it to be of the general type for the group — a 

 plated perisome traversed by ambulacra ; as might be expected in such small specimens this 

 is very delicate, the ventral side as shown at the inrolled portion of the arm (PL XI, fig. 15&) 

 being covered with the smallest imaginable plates ; the central part with the orals is not 

 preserved. 



The known stratigraphic range of Pycnosaccus has been increased by the discovery of a 

 species from the Keyser beds of West Virginia, belonging to the earliest Devonian ; and also 

 of an imperfectly known species from the Middle Devonian of the Mackenzie River basin 

 in Canada. 1 The genus is well distributed geographically, being found in Gotland, England 

 and America. P. costatus, founded by Angelin on a very small specimen which cannot be 

 found, is clearly a young P. scrobiculatus. Weller's imperfect P. americanus, from the 

 Niagaran beds at Chicago, is of the same type. Lecanocrinus calyculus, from the Niagaran 

 limestone of New York, proves to belong to this genus, as well as a somewhat similar speci- 

 men with unusually heavy plates from Chicago, figured by Weller as Lecanocrinus waukoma. 

 Arachnocrinus canadensis Whiteaves, from the Devonian of the Arctic region, should also be 

 placed here. Including these and P. dubius of doubtful generic position, there are 10 species, 

 which may be arranged as follows : 



The Species of Pycnosaccus 

 I. IBr usually 2. 



Crown usually small. 

 Calyx turbinate. 



Calyx angular with sharp ridges connecting centers of plates, 

 and pits at angles where the 3 sutures meet. 

 IBr moderately wide, arms round. 

 Base broad, but not projecting. 



Anal x broad and truncate P. scrobiculatus. 



Calyx ridges obscure. 



IBr wide, arms heavy. 



Anal x narrow and pointed P. nodulosus. 



Calyx smooth, plates without median elevation ; no calyx ridges. 



IBr wide P. calyculus. 



1 And it has been still further increased in the opposite direction by the acquisition too late for 

 -description herein of specimens of a new species from St. Paul, Indiana, thus carrying the genus back 

 to the Laurel formation. 



