228 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
early stages iu its formation. Tlie amount of forward extension will, however, be over- 
rated unless it be borne iu mind how much greater the ratio of the brain region to the 
trunk is in the earlier stages than in the later. 
Mesoderm . — The history of tlie mesoderm may here be conveniently carried up to ' 
the closure of the blastopore. The condition shown in Figs. 60, 61, 62, 63, 64, PI. ' 
xcv and xcvi (successive sections from same embryo, 29 hours), is practically retained 
until the closure. The mesoderm now throughout the trunk consists of two thick 
lateral masses (Figs. 60 and 61, mes.), thinning away at the sides, but not exhibiting the 
two-layered arrangement which in the trout i)refigures the ditferentiation of somato- 
pleure and splanchuopleure. As will be seen in the sequel, the development of the 
body cavity in the Bass is secondarily modified in a high degree. The cells of thef ( 
mesoderm plates are polygonal except at the surface, whei e they now begin to assume ) 
a columnar shape which later becomes more pronounced. “i' j 
The mesoderm plates fade away as they reach the neck region, here giving place' f 
to scattered mesoblast cells (Fig. 62). In the posterior head region there are a few ?' 
scattered ceils (Fig. 63) and a rather well-marked layer [en. mes.) which posteriorly is 
continuous with the entoderm, and anteriorly breaks up into scattered mesoblast cells, ,, 
which form a loose investment of the eye-balls and brain. The layer in question is 
no doubt in great part eutodermic, but it is impossible at this stage to fix upon the 
precise spot at which the entoderm lamella ceases and the scattered mesoderm begins. I 
All we can say is that in the anterior brain region (Fig. 64, PI. xovi) the primitive ' 
hypoblast becomes entirely converted into mesoblast. The anterior mass of meso- 
derm cells {a. ?n.. Fig. 55) disappears at about this time, its cells aiiparently migrating 
and sharing in the general investment of the brain. 
Formation of the mesoderm in Teleosts and the Coelom theory . — It has been univer- i 
sally recognized by students of the Teleosts that the mesoderm originates by delami- 
nation. It seems likewise certain that the primitive method of forming the mesoderm 
in the Chordata was by the outgrowth of paired sacs from the dorso-lateral walls ofj f 
the archenteron. According to Balfour’s idea of the coelom theory t' 
tradiction between the two facts. As is well known, he believed 
chians the mesoblast was delaminated as two lateral masses (more recent iuvestiga-j| 
tions would seem to show that it arises as paired outgrowths, Babl, 3S), but that theji 
delamination was the cceuogeuetic representative of an original outgrowth. He was]] 
thus thoroughly convinced that evaginatiou could be, and was, in certain vertebrates,] : 
replaced by delamination. | 
The coelom theory of Hertwig as applied to the vertebrates is, however, of a differ- j 
ent type. According to Hertwig not only has the method of forming the mesoderm in | 
the various vertebrate grouiis been derived from the primitive method shown iu -A-m'il 
phioxus, but in the ontogeny of all vertebrates it still follows the ancestral lines in] 
the one essential point : the mesoderm arises as paired outgrowths from the walls of { 
the archenteron. Hertwig has himself shown that the cavity of the sac may be oblit-,J 
crated and the outgrowth in consequence be solid, but he is convinced that the process! |' 
is always one of outgrowth and never of delaminatiou. His position is sharply defined|| 
in the following quotation: 
Bei keinem der Wirbelthiere entsteht der Mesoblast durch Abspaltung, sei es voiu iiussereii, sei 
es vom innereu, Grenzblatt, da er von beiden init Ausnaliine eines sebr bescbriinkten Keimbezirke8,'V 
uberall durch einem Spaltraum scliarf abgegrenzt wird (20, p. 308). 
