1908. | Innervation in Vaso-motor Reflexes, ete. 371 
briefly a possible hypothesis which unites what I may call the colloid- 
adsorption theory of Macdonald and the “ drainage” theories of von Uexkill 
and Macdougall with the synaptic membrane of Sherrington. The doubtful 
assumption made is the existence of membranes permeable to electrolytes 
in one direction only. As Overton has pointed out, as a physical structure 
such a membrane infringes the laws of thermo-dynamics, since it would, by 
Fig. 27.—Diagram of relations of general vaso-motor reflexes. A. Muscle-cell of arteriole. 
D. Vaso-dilator nerve-fibre terminating on A, and inhibiting it. C. Vaso- 
constrictor fibre also ending on A, but exciting it. These fibres arise in 
the dilator centre (D.C.) and the constrictor centre (C.C.) respectively. 
F. Depressor afferent fibre, dividing into two branches (or collaterals), one of 
which (—) inhibits the constrictor centre, while the other (+) excites the 
dilator centre. R. Pressor fibre of sensory nerve, causing rise of arterial 
pressure by exciting C.C. and inhibiting D.C. a, 6,¢,d. The respective 
synapses of these branches with the efferent neurones. The probable inter- 
mediate neurones are, for the sake of simplicity, omitted. 
creating potential differences, constitute a perpetual motion machine. I may 
venture to suggest that its existence may be ensured by the continuous 
supply of energy from the activity of cell-protoplasm. Macdonald has shown* 
that the colloids of the axis cylinder, when the nerve is injured, split off 
electrolytes, which were previously held in a “ masked ” state, or, as I prefer 
* ‘Roy. Soc. Proc.,’ vol. 76, B, p. 348, 1905. 
VOL. LXXX.—B. 2 & 
