27 
inwards by the outer wall of the dorsal hlood-chamher (Fig. 13, d.a.c.). The ventral air-chamher 
has corresponding relations (v.a.c.). In other words, in the special part of the gill there is a space 
between the general wall of the gill (composed of the chitine and hypoderm) and the boundary wall 
of the hlood-cavity. and this space contains air. This space does not communicate by any opening 
with the outside of the gill. 
I find that in many prei)arations of sections the hypoderm does not appear to be present 
as a layer lining the chitine in the special part of the gill (excepting at the free end of the gill). 
I am inclined to the opinion that the hypoderm may become Avithdrawn from the chitine and cen- 
tered in the pillars. As already observed, in this region of the gill the hypodermic nuclei are found 
only as associated with the pillars. It would seem possible that such a withdrawal of the hypo- 
derm may take place periodically with moulting as soon as a new chitinous layer has been foi’med. 
The hypodermic pillars in the special part of the gill are arranged in rows which divide 
the chambers into a series of compartments separated by broken walls. The appearance of these 
compartments, as seen Avhen the gill is examined from without has already been referred to, jn 47, 
as being a series of radiating blood channels communicating with the main excurrent channel. It 
will be understood, however, that the compartments of the dorsal blood-chamber only open directly 
into the main excurrent channel. 
Physiology of the outer gill. The course of the circulation of the blood. 
The course of the circulation of the blood in the outer gills of Oniscus can be seen by removing 
the first tAvo pairs of thoracic legs, placing the animal in water in a Avatch-glass, putting on a cover- 
glass, and examining under a low power of the microscope. Fig. 14 is based upon examinations 
thus made. The dotted arrows represent the courses taken by the in-flowing or venous blood and 
the lined arrows the courses taken by the out-flowing or arterial hlood. The blood, entering the 
gill at the basal joint, for the most part takes the path of the main incurrent channel. But as 
the inner boundary of this channel is a broken wall (formed by a row of hypodermic pillars) the 
hlood constantly escapes from the channel and spreads over the entire region of the general part 
of the gill. This distribution of the hlood further depends upon the curvature of the outer wall 
of the main in-current channel (the marginal waU of the gill) from which the current of blood is 
constantly being reflected. 
That portion of the blood which thus makes its way to the distal region of the gill then 
turns in its course and enters the main excurrent channel. The rest of the blood passes from the 
general to the special part of the gill, entering the radiating channels and proceeding directly out- 
wards toward the margin of the latter portion. Here it turns in its course and, folloAving the ra- 
diate channels but in the reverse direction, enters the main excurrent channel. A small portion 
of the blood, however, does not enter this channel but joins the outfloAving stream directly at the 
hasal joint. 
A portion of the blood, entering at the basal joint, does not go to the general part of the 
gill hut to the special part and appears to circulate as follows: It tends to spread over the region 
of the special part but mingling with the currents flowing from the general to the special part, 
its course is gradually lost in these currents. 
If now we apply the facts of structure and relation gained by the study of the anatomy 
of the gill to the explanation of the course of the circulation as observed from without, and above 
described, we are led to the following results: 
