ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. Pek) 
very fine imperforate thread. At its afferent and efferent extremities this connection 
is wider, hollow, and contains blood derived from the corresponding arteries. He there- 
fore regards it, and quite justifiably, as a ductus Botalli or vestigial aortic arch, corre- 
sponding to the one immediately in front of the existing first. In spite of numerous 
and most careful dissections of Myxine, I have not found any traces of this connection ; 
nor does Jackson find it in Bdellostoma dombeyi. He says:* “The last [ =my first] 
afferent branchial artery of each side gives off a small branch a short distance from the 
gill. This branch possesses a lumen only at its origin, if at all. It soon becomes reduced , 
to a slender string of connective tissue which becomes lost in the connective tissue 
around the ‘club muscle.’ Attached to this string is a small spheroidal body, apparently 
made up of fibrous and fatty tissue.” ‘In addition to the observation of MULLER, I 
have added that a spheroidal or flattened mass of connective tissue is found attached 
to the ‘ductus’ a short distance from its origin. ‘This body is larger and more saccular 
in appearance in Gdellostoma forsteri, and evidently may be interpreted as the rudiment 
of the gill pouch corresponding to the obliterated branchial artery.” The discovery of 
the rete mirable in Myxine naturally supports Jackson’s suggestion, as I have above 
indicated. 
The peculiar feature associated with the last afferent branchial artery occurs 
generally on the left side rather than on the right. The artery usually divides sooner, 
and forms two large vessels, the posterior of which gives off a branch which passes 
backwards and upwards, it may be for quite an appreciable distance ; but sooner or later 
it loses its lumen, becomes thread-like, and finally disappears altogether. Connected 
with it, as a rule, are several vascular papille (cp. fig. 1, tT). The significance of this 
structure becomes obvious when we investigate the variations in the gills. Of the eight 
eases I have carefully examined, six consisted of an extra gill on the left side only, 
and the remaining two of an extra gill on both sides. Also, in five out of these eight 
cases the extra gill was supplied by a branch from the last afferent branchial. Again, — 
in another case, with the normal number of gills, there were only five afferent branchial é 
arteries, the sixth gill being supplied by a branch from the last of these. I think, there- 
fore, that the blind twig from the last afferent branchial artery is associated with the 
former existence of a seventh pair of gills. 
The structures which I have called vascular papillee were first described in my pre- 
liminary paper. ‘hey are found both on the afferent and efferent branchial system 
(fig. 1), but are larger and more conspicuous on the former. They vary somewhat con- 
siderably both in number and structure, but I have never found them absent on a single | 
occasion, although a very large number of individuals have been examined. The reason 
why they have hitherto been overlooked on the afferent vessels (which are more usually 
dissected) is due, doubtless, to their position. As a rule, they are situated high up on 
the artery, near the point where it disappears into the gill, and are therefore tucked 
away under the etferent gill duct (e. g. d.); but they do on occasion occur lower down, 
* Univ. Cincinnati Bull., No. 5, 1901, pp. 21, 37. 
