ACTIVITIES OF COLONIAL ANIMALS 361 
then oozes through some of the siphonozodid pores, then dis- 
charges from the pores of the axial siphonozodid, as the india ink 
did, and finally escapes from the mouths of some of the auto- 
zooids. The conclusion to be drawn from these results is that 
the superior canal connects freely and directly with the opening 
of the axual siphonozo6id, as was first shown by Eisen (’76) and 
was subsequently confirmed by Wilson (’83),and that, contrary to 
my former opinion (719, p. 503), it likewise connects, though 
very much less freely, with the autozodids and siphonozodéids as 
maintained also by Wilson (’83). 
When a corresponding form of injection is carried out on the 
inferior canal of the peduncle, a solution of methylen blue yields 
exactly the same results as when it is injected into the superior 
eanal. India ink, however, remains limited to the inferior canal 
and only gradually makes its way out the mouths of some of 
the autozodids. It fails entirely to reach the siphonozodids or 
the axial pore. It is thus shown that the inferior canal in the 
rachis has freer communications with the autozodids than with 
the lateral or axial siphonozodids. 
That the superior and inferior canal systems communicate 
in the rachis is not only proved by the results given in the pre- 
ceding paragraphs, but may be demonstrated more directly in 
the following way. If an injection cannula is tied into the 
inferior canal of the peduncle in such a direction as to lead into 
the rachis and the superior canal is cut open from the peduncle to 
the axial pore, a methylen-blue solution on being injected into the 
inferior canal almost immediately appears in the open superior 
canal. If this procedure is carried out with india ink, instead 
of methylen blue, the ink fails to appear in the superior canal. 
The converse experiment of injecting into the superior canal and 
opening the inferior one yielded corresponding results except 
that there was often much loss of injection fluid at the axial pore. 
Both these lines of experimentation show that the superior and 
inferior canals in the body of the rachis are in communication 
with each other, but by openings not so large as those by which 
they communicate in the peduncle. 
