FOURTH REPORT OF THE DIRECTOR I907 163 



and illustrated.^ This attachment is so effected that the mouth of 

 the shell is seated directly over the anal aperture of the crinoid so 

 that the former may catch the digestal waste of the latter. Upon 

 this waste the Platyceras palpably sustains itself. So many 

 instances of this conjunction have passed under examination that 

 no question can arise as to the fact that such attachment is solely 

 for feeding purposes. Suggestions which have been occasionally 

 made that the attachment is rather accidental than otherwise, as 

 attachment to some substantial object is the habit of the gastropod, 

 are not borne out by the evidence afforded by multitudes of these 

 cases. It is quite certain, however, that in the Devonic and Carbonic 

 faunas where this habit became most prevalent, there was always a 

 predominant percentage of the gastropods that did not lend them- 

 selves to it ; nor have we reason yet to conclude that the habit once 

 inaugurated necessarily continued during the remaining life of the 

 individual. It did continue for a considerable period of the shell's 

 existence as the very instructive figure 6 on plate 6 indicates, the 

 concentric scars being the successive impressions of the lip of 

 the shell as its growth enlarged, while its position relative to the 

 after opening of the crinoid is unaltered. 



The history of this form of dependence is extraordinary and 

 illuminating. Throughout the Siluric the crinoids and C3^stids 

 abounded but moUusks of the limpetlike construction of Platy- 

 ceras were few. Moreover the crinoids were for the most part 

 built with slender domes well hedged about by delicate arms, and 

 on these domes the mollusk might have found difficulty in securing 

 a footing. 



The earliest intimation of the tendency on the part of a mollusk 

 to seek its food from the rejectamenta of the crinoid is afforded by 

 an example of a Lower Siluric Glyptocrinus which holds within its 

 arms and in a feeding posture a shell of the holostomatous gastro- 

 pod Cyclonema bilix. One might regard the occurrence 

 accidental if it had not been observed more than once. 



In the Upper Siluric, Platyceras had become somewhat more 

 abundant but its numerical development did not reach that of the 

 alHed mollusk Diaphorostoma and in plate 6, figure i , we have an illus- 

 tration of a small shell of this latter genus attached over the after 

 of the cystid Caryocrinus ornatus (Rochester shale). 

 Thus far in time no examples have come to our observation of 



]See particularly C. R. Keyes. Synopsis of American Carbonic Calyptracidae. Acad. Nat. 

 Sci. Phila. Proc. 1890. p. 150. The author here records a long list of these parasitic asso- 

 ciations and especially indicates the effect of this condition in modifying the aperture of 

 the gastropod. 



