508 JULIA B. PERATT. 
cross-section of but two or three cells above the dorsal wall of 
the spinal cord (figs. 8, 10, 11, 13). These cells appear to 
arise in the angle where the spinal cord separated from the 
skin, and a thickening in the ectoderm which forms the slight 
dorso-lateral ridge of the trunk seems occasionally about to 
add its deeper cell to the cells that have already migrated into 
the neural Anlage. Should this really take place, the dorso- 
lateral ridge of the trunk would be not potentially, but actually 
the homologue of the primitive dorso-lateral line of ectodermic 
proliferation on the head, of which it is the posterior continua- 
tion. Fig. 11 passes through such a cell above the fifth per- 
manent (i. e. myotome-forming) protovertebra. 
The dorsal ridge is very slight and soon disappears, but the 
median longitudinal ridge is well marked and continues com- 
paratively long. It is most conspicuous in those sections which 
pass through or near an intersegment, where the median longi- 
tudinal ridge unites with an intersegmental ridge. Figs. 9 and 
12 show the median longitudinal ridge between the fourth and 
fifth permanent protovertebre at two stages of development. 
Fig. 9 is from the younger embryo. It is seen that at an early 
stage the median ridge may be three cells deep near an inter- 
segment, while the surrounding ectoderm is composed of but 
two layers. In the later stage, given in fig. 12, the ridge is 
seen to extend to a point that reaches far in between the 
muscle plate and the pronephros. 
The appearance of the ridge as represented in figs. 9 and 12, 
though typical, is by no means constant. Neither is the ridge 
three cells deep at every intersegment in the younger embryo, 
nor does it extend so far into the lower tissues at each inter- 
segment of the older embryo. 
The boundary of the ectoderm, elsewhere true as if formed 
by a limiting membrane, often appears frayed on the ridge as 
if cells had just pulled away. Had they done so, however, it 
would be difficult to obtain positive evidence of the fact, for the 
reason that at this very time cells begin to wander towards the 
notochord from the dorsal extremity of the lateral plates, form- 
ing a scattered mesenchyme in which cells wandering from the 
