446 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
Their structure must therefore be the same at the time of segmeutatiou. We have 
here a differentiation due to external conditions similar to those found in the differ- 
entiation of sex from like cells by external conditions. While in infusoria this dif- 
ferentiation into macro and micro nuclear elements takes place after the division of a 
whole into two equal halves, the differentiation takes place before division in metazoa. 
Comparison beticeen the processes of conjugation in ciliate infusoria ( modified from 
Weismann after Maupas) and of maturation and segmentation in Cymatogaster aggre- 
gatus . — The modifications of the male cells are purely theoretical, and modified from 
the conditions often observed in a number of invertebrates. Nothing is yet known of 
spermogenesis of Cymatogaster. The black circles represent micro or germ nuclei, the 
blank circles the polar nuclei, the shaded parts the macronuclear elements. Note well 
that between stages H and A of protozoa there intervenes a large number of genera- 
tions of nuclei-individuals, and that a similar number of generations of nuclei, all of 
which collectively represent an individual, intervene between I and a = A in metazoa. 
In series I the macronuclear elements disappear in stage G, while in series in they 
do not disappear till much later than stage I. Usually in metazoa they disappear in 
stage A 2, i. e., before the nucleus from which they have been derived loses its entity. 
While in series i the macronuclear elements are segregated in stage F or at the begin- 
ning of the series of daughter nuclei, in series hi this process does not take place till 
stage A 2 is reached or the end of the series of daughter nuclei or concomitantly with 
the production of a new mother nucleus; in all the intervening stages between i and 
A 2 macro and micro nuclear elements are united in the same nucleus. 
Formation of the mesoderm. — Dr. Minot has said (Am. Nat., 1890, 877) : “Scarcely 
an embryologist can be found who has not published opinions on this question (origin 
of the mesoderm) considerably at variance with those of most authors.” To these 
already numerous accounts I must add that of Cymatogaster. Without question the 
mesoderm arises here from the entoderm, as has been observed in a number of other 
fishes by various authors. Instead, however, of being restricted to and arising from 
a narrow space, it is split off from the entoderm and forms a layer over the whole 
entoderm, exclusive of the axial line, or the region occupied by the chorda. 
Hertwig states, p. 119 : 
Bei kinem Wirbelthiere entstehen die Keimblatter durch Abspaltung, sei es vom auseren, sei es vom 
inneren Grenzblatt, da sie von beiden mit Ausnahme eines sehr beschrankten Keimbezirks iiberall 
durch einen Spaltraume scharf abgegrenzt werden. 
I have nowhere seen any figures in Cymatogaster which favor Hertwig’s view of the 
origin of the mesoderm, and since it apparently appears simultaneously over the whole 
entoderm exclusive of the middle line, and is closely applied to the eutoderm when 
it usually is widely separate from it, the only explanation tenable seems to be that the 
mesoderm is split off from the entoderm every where except at the median line. In all 
cases I have been able to examine the entoderm formed a layer beneath the chorda, but 
I am not positive whether a layer is always present beneath the chorda at the time the 
latter structure is differentiated, or whether the whole central portion of the entoderm 
is differentiated into notochord, and a new layer of entoderm is formed beneath this 
by ingrowth from the sides. The former seems the more probable view. 
The development of the chorda and of the mesoderm is still obscure, since I 
obtained but one or two marked stages between the closing of the blastopore and 
embryos with three proto vertebrae. On reconsideration 1 am not so certain as to what 
