THE NEMATOCYSTS OF EOLIDS 
137 
possible the easy passage of nematoc^^sts from the digestive tract 
to the cnidophores ; that specially adapted cnidophages store the 
nettlts, and become converted, with the aid of interstitial cells, 
into cnidocysts; tKat these step out of line and are forced by means 
of a special musculature out of the cnidopore; and that there is 
a zone of ^'embryonic" tissue which furnishes a supply of cells 
for all the purposes of the cnidophore. In addition to elimination 
there are other ways in which the overloading of the cerata is 
prevented, for new appendages are constantly forming, and by 
autotomy, the largest and presumably best stocked cnidophores 
are frequently cast upon the slightest stimulation. Even in the 
absence of assignable specific reasons for the voidance of nettles 
it seems reasonable, in face of these facts to assume that the pro- 
cesses now known to occur are useful, and that elimination of 
nematocysts is the primar}- function of the cnidophore proper. 
Grosvenor has given a possible history of this function and of 
the system of organs that carry it on. '^In molluscs," he writes, 
other than the cladohepatic Nudibranchs, the food is digested 
in the stomach, where absorption takes place ... In Tritonia, 
therefore, the anus suffices for the passage of nematocysts out of 
the body. But in the Cladohepatica, part of the food is digested 
in the gastric gland, quite fresh pieces of hydroid being found in 
the ducts and ceratal diverticula of a recently fed ^olid. . . 
How Dotonids which habitually feed on hydroids and have no 
aperatures in their cerata, get rid of the nematocysts, I cannot 
say; perhaps by throwing off their cerata, which as is well known 
they do with great ease. \\^en an aperture for the extrusion of 
nematocysts had once been acquired, it would be obviously advan- 
tageous that the distal end of the '4iepatic diverticulum" should 
be modified to form a cnidosac where the nematocysts might be 
stored." With this attempt at phylogenetic explanation, and the 
reasons cited to support it, I agree with one exception: to me 
it seems more probable that the ''cnidosacs" as well as the habit 
of storing nettles are both older than the cnidopore from which 
extrusion takes place, and furthermore, that primitively, eolids 
probably cast off their excess nettles by autotomizing the cerata. 
As a mechanism for voiding nematocysts, the cnidophoral 
