I24 THE AMERICAN NATURALIST. [Vor. XXXIV. 
the author urges (1) that in the Enteropneusta and Cephalo- 
chordata the number of gill slits is indefinite, new ones being 
added posteriorly throughout life; (2) that the number of gill 
slits has become reduced not only among the Vertebrata, but 
even within the group of the Enteropneusta, since one species, 
Ptychodera auriantiaca, has as many as seven hundred pairs, 
while another species, P. mzmeta, never has more than forty 
pairs, these representing the two extremes so far as known; 
(3) that in the intestinal pores and ciliated grooves of certain 
Enteropneusta we have vestiges of the former posterior gill 
slits of the now abbreviated series. The intestinal pores occur 
in six species belonging to four genera. These pores may lie 
either close behind the branchial region or at the posterior end 
of the hepatic region. Schimkewitsch and Spengel have sug- 
gested that these pores may in some way be related to gill slits. 
In the Ptychoderidz the pores are absent, but instead of them 
we have a pair of ciliated grooves extending from the anterior 
end of the hepatic region nearly to the posterior end of the 
body. In two species “these are not simple longitudinal fur- 
rows, but undergo metameric or interannular sacculations," the 
saccules * strongly resembling gill pouches” not yet open to 
the exterior Willey believes that the restriction of the gill 
slits to the anterior part of the body and of the gonads to the 
middle portion of the trunk took place after the development of 
a blood vascular system, which rendered the gonads no longer 
directly dependent upon the gill slits for oxygenation. 
Certain species'of Enteropneusta have two proboscis canals 
connecting anteriorly with the proboscis coelom and opening 
posteriorly by two pores on the dorsal side of the stalk of the 
proboscis, near where it joins the collar. The distal extremity 
of each of these canals may be swollen to form a considerable 
vesicle, or “end sac." Secondarily, in some species one (or 
both ?) of the proboscis canals may be interrupted, so that the 
end sac, while opening to the exterior, has no longer any com- 
munication with the proboscis coelom. In certain adults, though 
not in the young, of one species, Ptychodera carnosa, Willey found 
that one of the two proboscis pores ** may open ” not onto the 
