886 The American Naturalist. [October, 
duces the notochord and the mesoderm; the three processes 
going on simultaneously. The accompanying Fig. 24 represents 
a cross section of a larva with segments. The ectoderm, Ee, 
everywhere bounds the section; on 
the dorsal side a portion of the ecto- 
derm has been separated off to form 
the medullary plate, Md, above 
which isa small cavity. The cavity, 
In, of the archenteron is irregular, 
but symmetrical in outline ; the ento- 
derm bounding it can be separated 
into four parts: 1°, the lower por- 
tion, which forms the permanent en- 
FIG. 24.—Transverse section of an toderm, Ent , 2°, the upper median 
mphioxus embryo ; after Hatschek. portion, which becomes the noto- 
Meu e ps Lr chord, Ch; 3°, 4?, two lateral por- 
derm; Zn, archenteric cavity; Ms, tions, constituting the diverticula, 
mesodermic segments. Ms ; each diverticulum is a sepa- 
rate pouch, and as the development progresses, there are 
formed a series of pairs of pouches, stretching on either 
side along the notochord; later the pouches separate altogether 
" from the archenteron, each becoming a closed sack; the first 
pair of pouches, however, retain their connection for a con- 
siderable period with the archenteron, and have been described 
by older writers as glandular organs. The development of the 
pouches is, with the exception noted, most advanced anteriorly, 
and as we go tailwards the pouches are less and less advanced in 
development, until, as shown in Fig. 25, they merge into the 
general entoderm as a band of cells, Mes ; the last of which is the 
" mesoblast," Mb, a large granular cell, quite distinct from the 
remaining cells of the band or pouches. The pouches are the 
primitive segments (Ursegmente, mesoblastic somites of Balfour). 
In Amphioxus, then, the mesoderm arises from the entoderm 
along two lines, and is divided into paired hollow segments before 
it is separated from the mesoderm. Some writers, especially the 
brothers Hertwig, think this process of development to be primi- 
tive, and that the vertebrate type is derived from it. In true ver- 
