1922.) ** Renal Portal”’ System. 143 
with the blood system, thus assuming the function of a lymph 
which the restricted coelomic spaves have become secondarily 
continuous with the blood channels, and once in the lower 
Vertebrata, the kidney nephrostomes all disappearing when the 
coelom assumes the functions of a tit space (9). It is also 
significant that the lymph never has any communication with 
any space of whatever kind until this is completely shut off 
from the external world, instances of which, in addition to the 
coelom already named, are the brain-cavity and the cavities of 
the eye-ball and auditory sac ; also that in every part of the 
body at which the lymph does, on behalf of the organism part 
with water and at $5 the outside or a _ continuous wit 
the outside, the cells concerned are, with one exception! 
under strict nervous control, It follows then that if glomeru- 
lar filtration be a fact, it is the unique instance in the entire 
animal kingdom of blood freely (very pitins filtering off into 
a space continuous with the outside of the 
Another consideration is the very fondle fact that no 
other capillaries in the body are known to filter off the non- 
colloidal constituents of blood, leaving the colloids behind: 
even the most watery kind of lymph? contains 2% of proteids. 
Moreover one of the conditions of lvmph formation is the 
“ chemical ”’ influence exerted by the tissues traversed by the 
blood—an influence entirely absent in the case of the glomerular. 
capillaries: nor are the liver or limb ies under the 
necessity of forcing their exudate through any membrane 
resembling that ef the capsule—many such cr indeed 
effectually prevent fiyeeion (e.g. the lining epithelium of the 
frog’s lung, the membrane of Descemet covering the cornea. 
the lining oop ee of dinnd ducts and in all probability 
the outer wall of the glomerular capsule itself). 
Again, the development of such special structures as encap- 
sulated glomeruli —_— to produce water and salts is so 
ig superfluous when we remember that the lungs and 
in, the salivary pose the pancreas and the liver can all 

! The one exception is the liver, and in this case the cepa pe is not 
et seo — rial blood under high pressure but from v ood 
w pressure, and, mr - the 1 cee the quantity of Means is 
e 
pr as rr the contained ee of nce is from £35 8% Pro: 
essor Bayliss has however referred me to the paper by Scott (Jour. Physio- 
logy Vol. 50, 1915, p. 15.) in lara “it is “topes that fluid devoid of 
colloids, can re-enter the capillaries from the tissues and it would therefore 
appear to be possible for fluid, devoid of eullnide. to leave the cariilanien. 
