346 Prof. J. S. Macdonald. [Apr. 17, 
our theoretical strip with its alternate segments of high and low states 
of colloid solution, segments which we have now decided must be definitely 
divided off from one another by intersecting membranes, let us suppose 
that alkali is formed in the segments where the colloid particles are now 
larger than normal, that acid is formed in the segments where the state of 
more minute subdivision exists—and this as a secondary consequence of 
the existence of these states and of the changes in osmotic pressure so 
caused. It is conceivable that these chemical facts might be prominently 
useful in the return of the solutions, the one up to, the other down to its 
resting state. In other words, whereas contraction is probably a physical 
fact produced primarily by physical agency, relaxation must be (see above) 
to some extent a chemical fact and may be initiated by the chemical 
processes resulting from polarisation. 
Assuming this hypothesis to be correct, and muscular contraction dependent 
upon the excitation of a “ pseudo-polarisable ” colloid solution containing an 
electrolyte of the kind supposed, it is obvious that a strip of homogeneous 
colloid solution might become segmented in the manner of the muscle-fibril as. 
a consequence of its initial participation in function. Excite such a strip 
a small number of times in comparatively rapid succession. Arouse in it 
repeatedly this segmented arrangement of states of colloid solution, and it is 
to be expected that a permanent differentiation of this kind must ensue. The 
one segment is repeatedly affected in one direction, and brought back from that 
condition toa state of rest; its neighbour is repeatedly affected in the opposite 
direction and from thence returned to the same state of rest. This in fact 
seems so completely probable a suggestion, that one is inclined to lay stress 
upon it as a marked instance of the determination of structure by function. 
Considering the possibilities of the manner in which the function of the 
muscle might be served by this creation of conditions of high and low osmotic 
pressure in such alternate segments, the following points seem worth noticing. 
In the first place there is the resulting translation of water, to which it has 
not been found necessary to attract attention in the case of nerve-function, 
because of the absence of transverse membranes. The pressure in the dim 
bands being raised, and that in the bright bands lowered, water will pass into 
the dim bands from the neighbouring bright bands. If the lateral walls of the 
dim bands yield to the pressure ever so little, then these bands will assume 
the elliptical form, which they have been seen to adopt. The result of this 
process therefore is a movement of water from the longitudinal axis of the 
fibril, and thus from the longitudinal axis of the fibre, into a series of transverse 
axes. Thisis initself sufficient to account for the shortening and simultaneous 
thickening of the fibre as a whole. 
