No.1.] THE DEVELOPMENT OF BALANOGLOSSUS. 67 
invagination can be brought over the external openings of the 
gill-slits, so that they no longer can be seen in surface views 
of the young worm. This was seen to happen repeatedly in 
the living animals. 
The serial repetitions of the gill-slits of Balanoglossus, 
Ascidians and Amphioxus is interesting, although it does not 
necessarily indicate a phylogenetic connection. Each may 
have started with a single pair and pursued the same path 
independently. At present we cannot answer the question 
positively in either way. On the one hand we find in the most 
astonishing places external openings between the endoderm 
and ectoderm. In certain jelly-fish we find such communica- 
tions at regular intervals around the periphery of the bell. 
In the Mollusca, in Nudibranchs, we have regular openings at 
the ends of the dorsal papillae. In the Actinians, the Cinclides. 
Perhaps the anal opening belongs to this category. Similar 
fusions between the ectoderm and mesoderm are also to be 
found, as in the serially arranged dorsal-pores of certain 
Annelids. All this goes to show, I think, the possibility of 
independent origin of the se7zes of gill-slits. 
On the other side there stands one fact that seems to me of 
greater weight than the negative evidence of such general 
considerations. I refer to the presence and structure of the 
supporting-bars of the gill-slits. The structure of the support- 
ing-rods of Balanoglossus and Amphioxus shows an astonish- 
ing agreement. In each, chitin-like rods are present in the 
between-bars (primary bars) and the tongue-bars (secondary 
bars). In each the rods have a double origin, both for the 
primary and secondary bars. In each the rods are united 
along the dorsal edge of the gill-slits, but remain free below. 
Spaces that seem to be blood-vessels lie at the angles formed 
by the union of the rods. The body-cavities in both forms 
extend throughout the length of the bars. 
Certain minor points of difference are to be noted. Accord- 
ing to the figures given by Lankester the chitin-like rods unite 
with one another along the length of the bars, the union being 
closer in the tongue-bars. In the sfeczes of Balanoglossus 
described here only the rods in the between-bars unite into a 
