472 HaAroLD HEATH, 
heart is in part separated from the pericardial wall, and the ventral 
limbs of the coelomoducts are ensheathed in gland cells. These are 
the most obvious resemblances, and point to a community of descent, 
but beyond this point the question at the present time cannot be 
discussed with profit. 
If it is difficult to frame anything approaching a convincing 
argument regarding the relationships of one genus, the matter be- 
comes much more involved when the interrelationships of the entire 
order of solenogastres are concerned. With only a small number 
of genera known to the student of the group, and these usually 
clearly defined and not obviously related, most of the theorizing 
concerning their phylogenetic development must rest in large mea- 
sure on personal opinion, and therefore is usually convineing only 
to its author. The same is true also in the discussion regarding 
the most archaic solenogastre or the most primitive mollusc. Accor- 
ding to one investigator certain structures are declared to be primi- 
tive, and by another are considered to be coenogenetic. This is 
perhaps not a surprising state of affairs when we consider how far 
we are from possessing a broad and fundamental knowledge of the 
anatomy and embryology of molluses, but it is to be hoped that 
especially in this last named field we may have more data to support 
our personal beliefs. 
With the foregoing in mind I approach the subject of the primi- 
tive type of solenogastre in no dogmatie spirit, but after studying 
nearly fifty species of these animals certain facts have appeared 
that are suggestive. In another connection it has been argued 
(HeatH 1905), as other investigators have done, that the soleno- 
gastres appear to be more closely related to the chitons than to 
any other order. The symmetry of the body. the nature of the 
spines and their development, the snout and its innervation, the 
position of the heart, the relation of the aorta to the gonad and the 
head cavity, the relation of the pericardium and the coelomoducts, 
the two pairs of pedal glands and finally the nervous system are 
all fundamentally alike in the two groups. 
Externally the individuals of the two orders differ much in 
appearance, but whether the worm-like solenogastre is primitive in 
this respect or whether it, like Chitonellus, has assumed this shape 
secondarily is not known. Among a few embryos of Halomenia gra- 
vida in my possession are some advanced larvae that in form very 
closely resemble chiton embryos immediately before their meta- 
