The Evolution of the Kidney 55 



merular.^'^^'"'^*'^^ There is no constant rule by which the 

 aglomerular condition is reached; Grafflin^^ has shown that 

 in the "daddy" sculpin the glomeruU cease to function be- 

 tween the young and the adult stage, while Armstrong (per- 

 sonal communication) has shown that in the toadfish and 

 pipefish a glomerulus does not develop even in the embryo. 

 Though evolution is not reversible, the marine teleosts are 

 indirectly converting their kidneys back to the purely tubu- 

 lar form possessed by the prochordate ancestor which left the 

 sea in Cambrian or Ordovician time. 



But there is more than one way of solving physiological 

 difficulties, including that faced by the Ancient Mariner and 

 the marine teleosts. Let us return to the elasmobranchs, who 

 had first made the marine migration in the Devonian. These 

 more primitive fishes solved the problem of living in salt 

 water in an entirely different way. The four orders of the 

 sub-class Elasmobranchii — the sharks' rays, skates and chimae- 

 ras — separated from the parent stem and from each other in 

 or shortly after the Devonian period; that is to say, the De- 

 vonian is the most recent time at which we can assign to all 

 four orders a common ancestry. Yet all four orders possess 

 a common and surprisingly unique adaptation for living in 

 seawater; they have changed the composition of their blood 

 by deliberately bringing themselves, as it were, into a per- 

 petually uremic state; they reabsorb from the glomerular 

 filtrate as it passes down the tubules such urea as is present 

 in this fluid (urea being the chief product of nitrogen com- 

 bustion) much as the Ordovician-Silurian fishes learned to 

 reabsorb glucose and chloride. They return this otherwise 

 inert waste product to the blood until it reaches concentra- 

 tions of 2000 to 2 500 mgm. per cent. The presence of this 

 urea raises the osmotic pressure of the blood above that of the 



