652 THE SECRETION OF URINE. 
in diabetes insipidus, the urine secreted may have A = o- 16 C, i.e. a 
difference between A of blood and of urine of -4° C. Hence he con- 
cludes that the production of urine by the glomeruli is also attended 
with the doing of work, and must therefore be looked upon as a process 
of secretion. We might, however, still adhere to the theory of glomer- 
ular filtration, if we assumed either that the cells of the tubules could 
absorb water or solids according to the needs of the organism, or that 
they were able to secrete pure water and so dilute the glomerular 
transudation. 
Heidenhain's criticism of the theory of Ludwig. — The difficulties 
in the way of accepting Ludwig's hypothesis have led Heidenhain, after 
a long series of researches on the subject, to reject this theory abso- 
lutely, in favour of one very similar to that put forward by Bowman. 
Heidenhain sums up his objections to the mechanical theory under 
the following headings : — 
1. The hypothesis that a rise of arterial pressure causes increased 
transudation through the vessel walls, is not confirmed by our experience 
in other parts of the body (lymphatics of the limbs, salivary glands). 
2. This hypothesis is rendered the more improbable for the kidney, 
since in this situation the glomerular capillaries are covered by a second 
layer of epithelium, and we know, from Leber's researches on the cornea, 
that such a simple epithelial layer can afford great resistance to 
filtration. 
3. If we assume that all the constituents of the urine are filtered 
off in the glomeruli, the small amount of urea in the blood renders it 
necessary that in man about 70 kilos, of fluid should be filtered through 
and reabsorbed, in order that the urea produced in the course of 
the day may be excreted in the urine, — an amount which is highly 
improbable. 
4. According to the filtration hypothesis, the amount of urine formed 
must always increase with increased capillary pressure, whereas we find 
that, on increasing capillary pressure by ligature of the renal vein, the 
urinary flow is abolished. 
5. The hypothesis that the glomerular transudate is concentrated by 
a process of osmosis or diffusion, on its way through the glomeruli, is 
rendered impossible by the fact that the osmotic pressure of the 
urine may be, and generally is, much higher than that of the lymph 
or blood. 
6. The filtration hypothesis does not explain why the amount of 
urine is increased by the presence of water or crystalloid (Jiarnfahig) 
substances in the blood. 
From his own researches on the subject, Heidenhain conies to the 
following conclusions with regard to the mechanism of secretion : — 
1. In the kidney, as in all other glands, the secretion depends on the 
active intervention of special secretory cells. 
2. The first type of these cells is represented by the simple layer 
of epithelium covering the glomerular loop of capillaries. The office of 
these cells is to secrete water and such salts of the urine as are found 
in all other parts of the body in watery solution (e.g. sodium chloride). 
3. Another system of secretory cells, forming the lining investment 
of the convoluted tubules and ascending tubule of Henle, secrete the 
specific constituents of urine (urea, uric acid, etc.). Under some con- 
ditions they may at the same time secrete a certain amount of water. 
