448 
there are anterior to the same myotome 12 segments in the ventral 
series!). I would explain this discrepancy on the two sides of this 
individual by assuming a secondary subdivision of the ventral portion 
of one myotomic segment of the left side. That such a secondary 
subdivision of a myotome may take place in ontogeny I have positive 
evidence. For I find that in an embryo of Petr. marinus of 18 days 
(7!/, mm), the myosepta between the 6th and 7th and the 7th 
and 8th myotomes (which in this stage lie above the “Kiemenfurche”) 
divide at a point half way between the dorsal and ventral limit of the 
myosepta, so that each forms ventrally two myosepta. Thus a 
triangular muscle segment is seen between the 6th and 7th and 
between the 7th and 8th myotomes. Had such a secondary sub- 
division occurred in the myotomes behind the branchial region there 
would have resulted a discrepancy of two segments between the M. 
par. dorsalis and M. par. ventralis. And also in this individual, 
since such variation is seen only on the left side, a discrepancy of two 
segments in the M. par. ventralis of the two sides. 
The innervation of the M. par. subbranchialis throws still 
more light upon its developmental history. It is innervated, as has 
been shown by Kuprrer (95) by that nerve which he calls ramus 
recurrens vagi, and which he once considered a branch of the 
posterior ramus posttrematicus of the vagus. In a later paper (’96), 
however, he expresses a different opinion of its nature. He states 
(p. 617) that “es giebt bei Ammocoetes, wie Petromyzon, einen moto- 
rischen Nerven, welcher in seinem Verlauf Uebereinstimmung mit dem 
Hypoglossus der Gnathostomen zeigt. Es ist der aus dem Branchio- 
gastricus hervorgehende motorische R. recurrens vagi. Der Nerv um- 
greift, wie der Hypoglossus, die Kiemenregion hinten, gelangt an die 
Ventralseite und zieht dann nach vorn an der Innenfläche des ventralen 
Seitenmuskels, den er innervirt. Der Muskel ist aber, so weit ich es 
ermitteln konnte, ein Hautmuskel. Der Nerv besteht aus breiten 
Fasern, die im Branehiogastricus ein dorsales Bündel bilden und wahr- 
scheinlich aus den Anastomosen desselben mit ventralen Spinalnerven 
stammen.” How these anastomoses of spinal nerves with the “Epi- 
branchialstrang” of the vagus, first observed by Born (’27) and later 
1) It is to be noted that the number of segments of the ventral 
series in Petromyzon is greater than in Ammocoetes (assuming that the 
difference is not a specific one in the example given), as has been 
previously observed by WIEDERSHEIM (’80). The explanation for an increase 
in number at metamorphosis, if there be such, I am not able to give. 
