GAS GLANDS OF SOME TELdlOSTEAX FISHKS. 231 



not doubt.* But this explanation does not, in iny opinion, explain 

 all the facts, since, if mere slowing down of the blood current 

 is the one desideratum, then a rete solely connected with the 

 bladder artery is all-sufticient. Further, it is evident that the 

 explanation takes no account either of the fact that the bladder 

 vein also forms a rete mirabile or of the still more remarkable fact 

 that the retia on the bladder artery and vein respectively are both 

 formed at exactly the same distance from the gas gland and their 

 constituent arterioles and venules as intimately intermingled as any 

 product of human manufacture purposely so designed could be. The 

 suggestion Avhich I venture to ofter as to the use of this complex 

 arterial and venous rete universally associated Avitli gas glands 

 is the natural conclusion of the following considerations. The 

 fact that some hundreds of the finest capillaries conveying 

 blood to the gas gland are intimately intermingled with and 

 closely apposed to a like number (another significant fact) of 

 similar capillaries conveying blood from the gas gland is sug- 

 gestive of the exchange between the two sets of capillaries of some 

 substance necessarily of importance to the gas gland with which 

 the rete is connected. This hypothetical substance we must 

 assume both to be difi'used from the venous capillaries to the 

 arterial (since diffusion in an opposite direction would not aJBfect 

 the gas gland) and to be derived from the gas gland. The 

 question which next arises is why this hypothetical substance 

 should be poured into the arterial blood before it reaches the 

 gas gland, and the only answer that suggests itself is that 

 it is necessary for this substance to influence the arterial blood 

 in some manner during its passage (made slower by the rete) 

 from the rete to the gas glancl, so that by the time the arterial 

 blood reaches its destination, its constitution has become altered. 

 Now I have already stated that the gas gland, in the limited 

 sense of the term, is essentially an oxygen-producing gland and 

 that therefore the oxygen contained in the blood is the one 

 element which the gas gland requires, from which fact we may 

 conclude that the hypothetical substance referred to has some- 

 thing to do with the giving up of oxygen by the blood to the 

 gland. In the foregoing lines I have mentioned that such a 

 hypothetical substance has already been postulated by Jaeger to 

 explain the disintegration of the red blood corpuscles. We may 

 therefore state as extremely probable suppositions that in the rete 

 the venous capillaries contain a relatively large quantity of a 

 toxin poui"ed into the blood by the cells of the gas gland, that 

 this toxin diffuses from the venous capillaries into the arterial, 

 and that whilst the artei-ial blood is slowly travelling from the 



* This function, however, cannot he exercised in the case of the retia mirabilia of 

 the Eel, which are separated from the gas gland by a relatively few large vessels, 

 since the reunion of the arterial capillaries on the side next the sras gland means of 

 course the quickening of the blood stream. Neither can the Eel's rete mirabile 

 possess the supposed function of the " carotid gland " of Amphibia, of " deadening " 

 the heart pulse, since the gill capillary system of the fish must etfectively eliminate 

 all trace of this ; on the other hand, the rete mirabile must certainly minimize the 

 pressure of the blood supplying the gas gland, a result which, in the present instance, 

 we must regard as a defect. 



16* 



