28 
The blood enters the ventral blood-chamber and spreads ovit through the entire cavity. 
In the distal part of the gill where the partition separating the ventral from the dorsal chamber 
comes to an end the blood passes into the dorsal blood-chamher and immediately into that portion 
of it forming the main excurrent channel. The blood entering the ventral chamber of the special 
part of the gill flows through the compartments of that chamber to the places where they commu- 
nicate with the corresponding compartments of the dorsal chaml)er and there enters the latter. It 
flows through these to the main excurrent channel, excepting a small portion which passes into 
the excurrent channel at the basal joint. 
The function of the b lo o d- ch amb er s. The function of the ventral chamber is ob- 
viously to receive venous blood from the body and distribute it over the entire outer face of the 
gill. It is there separated from the external atmosphere by its own Avail and the ventral Avail of 
the gill. As this wall of the gill is rather thin it Avould appear that an exchange of gases may 
take place betAveen the blood in the ventral chamber and the external air. (This statement is in- 
tended to apply here only to the general part of the gill ; the exchange of gases in the special part 
is considered l)eloAv.) The function of the dorsal chamber is evidently to expose the blood still 
further to the action of air (as described heloAv) and finally to collect the purified blood into a 
channel through which it passes back to the body. The use of the middle chamber appears to be 
passive rather than active. The space taken up by it reduces by so much the depth of the 
chambers above and below and thus secures the exposime of a very thin stream of blood to the 
action of the air in the air-chambers. 
The function of the air-chambers. When the blood enters the ventral chamber of 
the special part of the gill it is separated from the air in tlie corresponding air chamber only by 
tlie single thin membrane which forms the outer boundary Avail of the blood-chamber. The ordinary 
conditions of the respiration of air are thus satisfied. As the blood passes at the margin of the 
gill into the dorsal chamber it is further exposed to the action of air in the dorsal air-chamber 
and the respiratory process is here completed. 
The air-chamhei’s, as already noted, do not communicate by 02 Aenings Avitli the exterior of 
the gill. The contained air must therefore enter by penetration through the wall of the gill. The 
wall, as Ave have seen, is here very thin, the hypoderm, as a rule, appearing to be withdrawn and 
leaving only a very thin layer of chitine. 
As has already been stated, when an animal is placed in water the apjAearance of air in 
the gills passes away. This observation led to a number of experiments the chief pimpose of which 
was to gain eAudence, in addition to the evidence derived from direct inspection and from the study 
of the structure of the gill, that air is normally present in the chambers.^ 
1. An animal Avas placed in water until the apiAearance of air in the gills had passed away. 
It Avas then killed and the tissues fixed by hot 83”/o alcohol. In sections of the gills pre|)ared 
from this specimen the air-chambers did not appear empty (as usual) but contained blood without 
corpuscles, that is, blood plasma (appearing as granulated matter). 
This experiment having indicated that under the conditions imposed the air was replaced, 
in part, at least by jAlasina of l)lood, we next sought to modify the experiment in order to ascertain 
whether the presence of blood in the chambers was due entirely to the conditions being artificial. 
‘ These experiments were suggested to me by Dr. ZUB Stbassen. 
