124 HarozD HEATH, 
buceal cirri, buccal ridges (Mundleisten), and buccal plate (Mund- 
schild) is much more involved. 
With the radula the case is different. In the present species 
it has been remarkably conservative, retaining its basement membrane 
and developing from sets of cells which, as has been shown previ- 
ously, are strikingly like what exists in the chitons. WırEn (1892) 
and THieze (1902) have maintained that a typical radula does not 
exist among species known to them for not only is the radula small 
but the teeth are exceedingly minute, without a basement membrane 
and seemingly developed from cells without any homolog outside of 
the group. Nevertheless Wrren himself has called attention to the 
close approximation of the radula in Proneomenia acuminata to this 
structure in the typical glossophora and HEusCHER (1892) has 
emphatically done the same for Pr. sluiteri. It is possible that a 
more attentive examination of certain species may disclose the 
presence of a basement membrane though it is an undoubted fact 
that the radula is in a degenerate condition in some species and is 
wholly wanting in others. In a species from Hawaii belonging to 
genus Myzomenia so far as my limited study permits me to judge 
there is a distichous radula which in its fully developed state seems 
to lack a basement membrane. However in the bottom of the radula 
sac where the teeth develop normally one is clearly present but it 
soon becomes so closely applied to the teeth as to become almost 
invisible and the difficulty of recognizing it is increased by the fact 
that where the radula opens into the mouth the membrane splits 
along the mid-line as in Z. talpordeus. 
The presence of a typical radula in Limifossor talpoideus (and 
in another undescribed species of the same genus) and of a subradular 
organ characteristically located and innervated points unmistakably 
to the fact that these organs existed in the ancestral form which 
in this and several other important respects was related to the 
chitons and prosobranchs. 
The musculature of the solenogastres and other molluscs does 
not appear to be a system upon which much dependence can be 
placed in determining the broader relationships of the larger groups. 
In the present species for example there is certainly no close 
thoroughgoing resemblance to what we find in Ch. nitidulum and 
the similarity to the Neomenüdae, judging from the study of 
Proneomenia hawaiiensis and Neomenia vampyrella, is not much closer. 
The same is true when we extend the comparison to the complex 
