126 
both the naso-pharyngeal (hypophysial) duct and the mouth cavity of 
Myxinoids are parts of an originally single cavity which later becomes 
divided into these two portions by the formation of a secondary 
horizontal partition, whilst in Petromyzontes the hypophysis is formed 
quite independently of the mouth cavity (Donrn 8, GörTTE 16, 17, 
Scorr 34). Since then it is almost certain that even the mouth 
cavities of Marsipobranchs and Gnathostomes are not equivalent 
structures we have still less reason to assume that the skeletal ele- 
ments associated with them must be homologous. 
Taking into account the whole of the evidence stated under the 
above five headings, I do not think it is possible to maintain any 
longer the hypothesis that Marsipobranchs are modified Gnathostomes. 
The desire to detect jaws in the Marsipobranch head is on a par 
with the desire to identify eircumanal skin ridges with pelvic fins, 
and both are expressions of that attitude of mind which, impressed 
with the antiquity of most of the characteristic features of organi- 
sation of well-defined phyla, views with suspicion any assertion that 
the absence of any characteristic features is other than a case of 
secondary disappearance or modification. That this attitude of suspicion 
is sometimes unjustifiable however is proved in the case of the group 
of animals under consideration, because most authors are agreed that 
many traits in which the Marsipobranchs might be supposed to be 
degenerate are rather to be regarded as extremely primitive. Apart 
from such primitive features as those already mentioned—the absence 
of jaws and the other visceral arches, the absence of limbs, the 
exceedingly simple condition of the wholly or partly membranous 
eranium (absence of occipital region, the 9th and 10th nerves passing 
out behind the cranium), the presence of lateral myotomes in the 
head—I may, among many others, just enumerate the following :— 
the position of the pituitary body, the persistence of all the metaotic 
myotomes, the absence of a horizontal septum dividing the myotomes 
into epiaxial and hypaxial portions and the junction of the myotomes 
in the mid-ventral line, the third pre-otic myotome (abducens muscle) 
is innervated on a plan different from that of Gnathostomes (Für- 
BRINGER, 12, JOHNSToN, 22) and in Myxinoidei all the three pre-otic 
myotomes have apparently never become specialized to form ocular 
muscles (judging from the development—von Kuprrer 25—and the 
primitive condition of the eyes—AtuEn t, DEAN 7, SEDGWICK 35), 
there is no articulation between the head and the simple non-verte- 
