DEVELOPMENT OE THE SKULL IN THE SALMON. 
117 
seen to be the same with those that succeed. These arches are formed in that part of 
the blastoderm which lies within the huge opening of the “ umbilical vesicle that 
vesicle attaches the fore part of its neck, at present, to the under surface of the rudi- 
mentary tongue, with its enclosed “ basihyal” cartilage (Plate I. tigs. 3 & 8, u.v., g.h ., 
v. 4), and because of the heart (A.), which is situate in the mid line, these bars are wide 
apart at present. There are five pairs of rods behind the hyoid arch ; these become 
the branchial arches {hr.), all save the last pair, which become dentigerous and are always 
feeble ; they all suffer from the general non-symmetry of the head, are all smaller than 
those in front of them, but all except the last early acquire the elegant sigmoid 
in-hooked form. That these Salmon-embryos rapidly assume the more highly specialized 
characters of a noble “ Teleostean” Fish, and are much further from the lower “ Myxinoid” 
types than the Frog-embryo, is evident if we compare my very earliest condition (Plate I. 
figs. 1 & 2) with the counterpart stage in the Frog (“Frog’s Skull,” Plate hi. fig. 3). 
That part of the primordial skull of the Salmon which is quasi-vertebral (its 
“ peri-notochordal” region) is very small, and forms the floor for the “ medulla oblon- 
gata” (Plate I. figs. 2 & 8, and Plate II. fig. 2, nc ., i.v.). Of all the facial arches the 
trabecuhe come nearest to the investing mass in this type (Plate I. figs. 2 & 5) ; whilst in 
the Frog (op. cit. Plate III. figs. 4-6) they are the furthest from it. In a number of 
specimens I was able to make out, most clearly, the free truncated ends of the moieties of 
the “ investing mass” or “ basilar plate” (Plate I. figs. 2 & 3, and Plate II. figs. 4 & 5, iv.). 
The abrupt manner in which this “puYe-pr 0 tovertebral ’ tract ends in front is quite 
remarkable; it is as square as though it were really one of the proto-vertebrae, the 
foremost*. 
In my most immature specimen (Plate I. figs. 1 & 2), studied as a transparency, but 
in the upper view (fig. 2), having the skull-floor partly laid bare from above, the twisted 
condition of the embryo gives the notochord and its investment a turn toward the left 
side ; it is seen to reach the trabeculae ; but its investing moieties (iv.) quite run short, 
so that there is a peculiar and sudden abortion of the primordial vertebral structure. 
All that is built upon this foundation is the occipital arch and the “ prootic bridge,” 
and the mansion of the brain is built up from many a source foreign to these mere axial 
rudiments. 
There is a skull-rudiment which develops in a manner unexpected by me ; this is the 
* I can now throw a little more light upon the Eowl’s skull than when I wrote upon it. Professor Huxley 
gives a figure of the primordial skull of the Fowl in his ‘Elements’ (p. 138, fig. 57, F 1 ), and reproduces it in 
his ‘Manual’ (p. 18), which agrees very closely with my first stage (Plate lxxxi. fig. 2). In these figures 
the “ trabeculae” converge towards each other over the ends of the investing mass, and this mass terminates 
in a truncated manner on each side, the squared end looking forwards and outwards ; the two structures are 
not seen, however, as distinct. In my next stage (Plate lxxxi. fig. 8) the blades of the trabecular forceps 
have opened out, and the investing mass is now mesiad of the apices of the trabeculae, which are plainly seen. 
Not so plainly seen, however, as in the third stage (Plate lxxxii. figs. 1, 2, 3, lg.), where these apices have 
become long, free, rounded rods, with a cellular interior, such as I find in the trabeculae of the embryo of the 
Eoa, Eunectes murinus. 
