THE FOSSIL CRINOID GENUS DOLATOCRHSTUS AND ITS ALLIES. 53 



species are founded. The testing of the value of such characters in 

 practice emphasizes the importance of ample material. I have 

 before me the types of 15 of Miller and Gurley's species, and in addi- 

 tion 31 duplicates in the Gurley collection, and 76 good specimens 

 of my own of which 62 show the full number of arms, thus making 

 a total of 108 specimens of this form from which the number of 

 arms can be tabulated, to which may be added the 9 types of Rowley. 

 These range from 10 arms to 22, distributed as follows: With 10 

 arms, 2; with 11 arms, 1; with 13 arms, 2; with 14 arms, 4; with 

 15 arms, 9; with 16 arms, 21; with 17 arms, 16; with 18 arms, 25; 

 with 19 arms, 16; with 20 arms, 18; with 21 arms, 1; with 22 arms, 2. 



While this shows a wide range of disturbance in the arm develop- 

 ment, it will be seen that the variation is chiefly within definite 

 limits, about 90 per cent of the specimens having from 15 to 20 arms, 

 and 50 per cent are within one or two of the higher number ; so that 

 normally this ionxi has either three or four arms to the ray, the 

 differences being due to the addition or failure of an arm irregularly 

 in one or more rays. 



These various numbers do not correlate with any other character; 

 and with the actual specimens in hand every attempt to arrange 

 species with reference to them is soon seen to be futile. The few 

 groups which I have tried to define according to other criteria all 

 embrace more or less of these variations in number of arms. As is 

 to be expected, these groups shade into one another in a most per- 

 plexing way; they are the outcome of frequent shif tings of specimens 

 from, one to another, without achieving any result that seems to 

 proclaim their identity by any positive or well-defined and palpable 

 character. This lack of reliable definition shows the small taxonomic 

 value of the variations noted, and it is probable that a far more drastic 

 cutting down of species, even to the extent of throwing them all into 

 one, would be the more logical course. In view of the considerable 

 reduction of species which I have felt obliged to make, and in order 

 to avoid too much dependence upon my unsupported judgment, I 

 am publishing herewith new and accurate drawings made from a 

 photographic basis of nearly all of Miller and Gurley's types, ar- 

 ranged in juxtaposition, so that the reader may judge for himself 

 (pis. 13, 14, and 15). These type specimens are in some cases not 

 so clear in point of distinctness of characters as could now be selected 

 for illustration; but I have not undertaken to prepare other figm'es, 

 beyond one or two to give a better view of the general type, to 

 illustrate variations, and to show that the purse-like contour which 

 is the striking character of this form is not due to age, since it appears 

 as distinctly in the smaller specimens as in the most mature (pi. 16). 



