IO SMITHSONIAN INSTITUTION 



panying remains does not conflict, as this would only mean that when the 

 crinoids perished any moderate current would carry away the crowns and 

 stems, leaving the large roots more or less imbedded in the mud where they 

 grew, as already explained. 



There was a tendency among some Paleozoic crinoids to enlargement of 

 the root, by which the terminal stem-branches were connected by a secondary 

 growth of calcareous matter, sometimes forming a mere flat-bottomed encrust- 

 ing mass adherent to other objects, and sometimes a complicated discoid body, 

 full of chambers or channels, resting in the mud. Such, for example, was the 

 Astroporites ottawensis described by Dr. Lambe * from the Trenton of Canada, 

 which I have found both attached to the rock and bedded in soft shale with the 

 lower surface undifferentiated. Dr. Sardeson * has described various forms 

 of such discoid roots, to which he has given the name Podolithus; and of which 

 he considers Camarocrinus to be a modified form. He traces its probable 

 origin by the aid of an interesting series of figures, supposing the spheroidal 

 mass to have ultimately become free, and to have floated in an inverted position. 



In another case the terminal stem-branches form a rounded mass, with 

 lateral, hook-like projections which would admirably serve the purpose of a 

 grapnel or anchor; hence these bodies have been named Ancyrocrinus.' But 

 no one knows to what crinoid they belong, as, though occurring quite abun- 

 dantly in the Hamilton beds at the Falls of the Ohio and elsewhere, no calyx 

 has yet been found associated with them. The statement sometimes made 

 that they belong to MyrtiUocrinus is without the slightest authority. A more 

 plausible supposition would be that they belong to Arachnocrinns , which has 

 a similar quadripartite canal, and one species of which occurs in the same 

 horizon. 4 Ancyrocrinus is found in the Hamilton of Clark County, Indiana; 

 Erie County, New York, and at Thedford, Ontario; but no authentic case of 

 it in the Onondaga. MyrtiUocrinus occurs only in the Onondaga; Arachno- 

 crinns also occurs in the Onondaga of New York, and chiefly in that formation 

 at Louisville, Kentucky; but one species — very rare — occurs in the Hamilton 

 at the latter locality. Except this last, no calyx having a quadripartite stem is 

 known from the Hamilton of this country. 



Other singular modifications in the distal growth of the stem are known 

 among fossil crinoids. Aspidocrinus Hall/ a circular disk with radiating ribs 



1 Canadian Record of Science, Jan.-Apr., p. 287, 1896. Also F. Springer, Geol. Surv. Canada, 

 mem. 15-P (Trenton Echinoderm Fauna at Kirkfield, Ont.), p. 46, 191 1. 



"Journal of Geology, vol. 16, pp. 239-254, 1898. 



"Hall, 15th Rep. New York State Cabinet (1862), p. 117. 



'Springer (New American Fossil Crinoids), Mem. Mus. Comp. Zool. Harvard, vol. 25, no. 3, 

 p. 122, 191 1. 



* Nat. Hist. New York, pt. 6, Pal., vol. 3, p. 122, pi. 5, figs. 15-20, 1859. 



