122 
ME. W. K. PAEKER ON THE STRUCTURE AND 
with the most lucid appearance of the relation of the investing mass to the trabeculse. 
These translucent Salmon-embryos show their parts with intense clearness, and here espe- 
cially a definite image is important. This individual had broader trabeculse (fig. 5, tr .) 
and investing mass with narrower ends than others examined ; the knee-shaped trabecular 
apices, bent before the pituitary body was within shot-range of them, turned inwards, 
and lay crosswise upon the free ends of the investing mass, overlapping them uncon- 
formably. 
The blades of these minute “ forceps ” evidently move, opening as they grow, and 
bring themselves into conformity to and into coalescence with the ends of the investing- 
mass ; this will be seen in the next stage. 
Third Stage . — Tlnliatched Salmon with Meckel’s cartilages free. 
This stage does not yield in interest to the last, and the growth now must be very 
rapid, the metamorphic changes taking place as fast as the development of flowers in 
spring time. 
I think it is very probable that the third stage is the morphological equivalent of that 
which is persistent in the Rays, just as in the unhatched chick we come upon “Rhyncho- 
saurian,” Struthiine, and Tinamine stages, as we delve from above downwards into the 
more simple and unspecialized conditions of the embryo Fowl. Certainly I shall soon 
come to a most evident “ Polypterine ” stage. Standing on the level of those depressed 
outspread “ Elasmobranchs,” the Rays, we see what fitnesses for aquatic life may arise 
in organisms halfway between the simplest possible conditions of a Fish and the noblest 
spiny-finned “Percoids.” If these fiat-headed, flat-bodied, long-tailed, broacl-flippered 
Salmon-embryos had grown largely without zoological improvement , to the Plagiostomes 
they must have gone, no other place would have been found for them *. 
One of the first preparations made by me of the Salmon-embryos was that which is figured 
in Plate II. figs. 6 & 7. This preparation of the primordial skull was very exquisite, but 
showed such strange characters that it had to wait for some weeks for interpretation ; 
the second stage, just described, was the sine qua non for that. Articulated to the fore 
and under part of the well-developed periotic capsule, I found two similar, nearly equal, 
strong cartilaginous rods, each growing athwart, under the head, and only directed 
slightly forwards. The foremost was too far hack for it to be the mandibular pier, and 
they both were too well developed, too large, and too far forwards and outwards to 
answer to the first and second branchials. Thus in a few moments I was able to reason 
out what they were not ; the missing link (my second stage, fig. 3) turning up, it soon 
became evident what they were. Yet this temporary conversion of the second “ postoral 
into a “double couple” of arches disturbed the complacent manner in which I was making 
the subdivision of the Tadpole’s facial parts into a measure for the rest ; the structures 
seen in that type are not the most perfect, but are low in kind, and therefore cannot be 
* Those of my earliest stage, in which the non-symmetry is at its least degree (Plate I. fig. 10, transverse 
section ), have the Raiine type of head in perfection. 
