646 
complete figures there is no indication whatever of a rectus oculi 
internus muscle. I have also not been able to find that muscle men- 
tioned in the text, but itis possible that I have there overlooked it. 
On pages 1252 and 1364 of GreIL’s work, the nervus oculomotorius 
is described in embryos of stages 47 and 48 respectively, and it is 
there said to first send a branch to the rectus superior, to then cross 
under the nervus ophthalmicus profundus, passing along the anterior 
surface of the rectus externus (lateralis, GREIL), and then to first supply 
the rectus inferior and then the obliquus inferior, ending in this 
latter muscle. No mention is here made of a rectus internus, and 
if that muscle is wanting there are but five eyemuscles in the fish, 
two obliqui and three recti. Brine and BurckHArpr (1905, p. 522) 
also say that there are but five eyemuscles in this fish, but they find 
four recti and but one obliquus, the obliquus superior. Van WisHE 
(1882) says that he found, in the adult of this same fish, the nervus 
oculomotorius innervating the “usual muscles”, which must mean 
four recti and two obliqui, the manner of innervation not being given. 
Whether there is an error in one or the other of these several des- 
criptions, or whether the explanation is that the eyemuscles in Ceratodus 
are variable, is an open question. GREIL’s figures of his models 
would certainly seem to be correct for the specimens he examined, 
for even in the text figures of transverse sections there is no slightest 
indication, that I can find, of a rectus internus muscle. According 
to Hyrrr (1845) both obliqui muscles are wanting in Lepidosiren, 
while according to Pınkus (1894) there are six muscles in Proto- 
pterus, and they are innervated as in selachians. 
Accepting GrEIL’s figures as correct and the conditions there 
shown as uniform in this fish, this arrangement of five muscles in- 
stead of six must either represent a retrograde condition that is inter- 
mediate between that found in selachians and those found in ganoids 
and teleosts (ALLIs, 1897 and 1903), or it must represent a primary 
condition from which all three of those arrangements have been derived. 
And as six eyemuscles are so generally, if not invariably, found in 
all other vertebrates in which the eyes are well developed, this number 
of muscles thus evidently being of advantage to the animal, it would 
seem as if the arrangement in Ceratodus could not have been derived 
by direct descent from an ancestor with six eyemuscles. 
Assuming such to be the case, the selachian arrangement would 
arise by the differentiation of a rectus internus from the rectus 
