Larva of Ohaoborus Grystallinus (de Geer). 365 
to the amount of air held in solution in water, indicating that a 
large volume of water has to be dealt with to obtain a limited 
supply of air. 
“ It is true that dissolved air is richer in oxygen than atmo- 
spheric air, about one-third of its volume consisting of this element, 
but the essential point to bear in mind is that the total volume 
of air held in solution in water at ordinary temperatures is so 
exceedingly small that in a litre of water the maximum amount 
of oxygen present is 10 cc.m., as compared with more than 
200 cc.m. in a litre of atmospheric air.” (Regnard P., 1891.) 
If the young larvae can be confined in a strictly limited amount 
of water, under a cover-glass, and not in contact with atmospheric 
air, the evidence is conclusive that it is itself a gas producer from 
its earliest stage, and the question naturally arises as to its 
remaining such during the whole of its larval existence. 
Our first examination of the air sacs, dissected out from 
the larva, disclosed the difference in the attachments at the 
two ends, one showing only a tracheal tube, the other a knob- 
shaped valve — already referred to. What is the purpose of this 
difference ? 
Repeated dissections, and taking out of the thoracic sacs, 
resulted in repeated failures to obtain more than very short lengths 
of the tracheal tube, or the tube attached to the valve, the closely 
packed muscles and other tissues tearing the tubes away whilst 
attempting to disengage the sacs. 
We ultimately tried the posterior pair, and finally succeeded 
in disentangling the two sacs with long transparent tubes attached 
to the valve ends. All four sacs are similarly equipped. We 
cannot say whether the tubes as extracted are complete, or whether 
some portions were detached in withdrawing them. 
The figure we give exactly represents the pair of tubes and 
sacs as displayed after dissection (fig. 32). 
The entire length of each tube is filled with cells, each having 
a clearly defined nucleus, and some of them appear to be 
vacuolated, the interstices between the cells being occupied by 
highly refractive granules. With some diffidence we put forward 
the hypothesis that the cells in these tubes are mainly respon- 
sible for secreting gas for the sacs, the latter containing both 
nitrogen and oxygen. This indicates two kinds of cells possessing 
the power of selection, similar to the cells in the mammalian 
kidney. On the other hand, the cells may secrete nitrogen only, 
the pigment spots on the air sacs helping in the process of 
respiration — as hinted by Miall, but he does not suggest how this 
could be brought about. 
We fail to see under these latter conditions how the propor- 
tions of the two gases could be maintained ; also it implies the sac 
is pervious to air, its collection and distribution taking place 
