527 
find well developed in the eye of the chain Cyclosalpa pinnata, and more 
rudimentary in the chain forms of other Cyclosalpae, a small compact 
mass of optic cells, lying in the hollow of the horseshoe and ap- 
parently somewhat distinct from the latter. These conditions do not 
indicate metamerization, nor can 1 see in them any significance in 
connection with the phylogeny of the eyes of vertebrates. 
To one familiar with the development of the eye of Salpa any such 
phylogenetic significance seems impossible. The primary eye of the 
Salpidae is a simple, continuous, horseshoe-shaped mass of optic cells 
with no suggestion of any metamerization or division into bilaterally 
paired portions, This eye is found, practically unmodified, in the 
solitary individuals of all species. The eye of the chain Salpae is 
modified, having been derived from an eye like that of the solitary 
form by a somewhat complicated process of inversion, by the suppres- 
sion of certain portions in many species, and (in the Cyclosalpae) by 
the greater development of its primitively posterior end. All the con- 
ditions are secondary and derivative in the eyes of tbe chain indivi- 
duals, and we find here the greatest variety of special conditions 
characteristic of the several species. 
If, therefore, any indications as to the phylogeny of the eyes of 
vertebrates were to be found in the eyes of Salpidae, they should be 
sought in the eyes of the solitary salpae rather than in the secondary 
and greatly modified eyes of the chain individuals. 
But the eye even of the solitary Salpa has little light to throw 
upon the history of the eyes of vertebrates. The whole nervous system 
of Salpa is highly modified. In it is retained only one of the three 
portions into which the central nervous system of tunicate larvae is 
divided. The central nervous system of the tunicate tadpole has a 
sensory vesicle, a trunk portion, and a caudal portion. The whole 
ganglion of Salpa corresponds to the adult ascidian ganglion and 
neural gland and to only the trunk portion of the ascidian tadpole’s 
nerve tube. 
This ganglion in Salpa is greatly enlarged by a secondary pro- 
liferation of cells during the early development, and only rather late 
in the development does there appear on the top of this secondary 
mass of cells a group of optic cells with an arrangement utterly dif- 
ferent from that of the little eye of the ascidian tadpole. It is as 
unwarranted to attempt to draw from the condition of the ganglion 
or eye of Salpa conclusions as to the phylogeny of the organs of verte- 
brates, as it would be to draw similar conclusions from the condition 
of the organs in the adult ascidian. It is in the ascidian tadpole and 
