bright bands is drawn into the dim bands between the short rods, or sarcostyles, which are consequently separated more widely. No tissue could be more unsuitable than muscle for microscopic examination; for none other offers the same optical difficulties. This will be evident to anyone who considers the description already given of the markings which it exhibits. Whatever may be the true interpretation of these markings, it is clear that they point to an almost infinite multiplication of minute elements adjusted with absolute accuracy side by side and end to end. A cylinder filled with these transparent objects has to be viewed by transmitted light. The elements, whatever may be their nature, refract light in different degrees. It is impossible to eliminate the effects of internal reflection, refraction, and interference of waves of light. The most alluring hypothesis must be accepted with a considerable amount of reserve. Any fact which seems to militate against it must be taken into consideration. The view set forth above, in general terms, is very attractive to everyone who wishes to bring muscle within the category of machines. Suppose we accept the hypothesis that the dim band is a plate made of sarcostyles surrounded by sarcoplasm then the impulse which reaches a fibre causes an alteration in the surface relations of the rods to the substance in which they are embedded. Molecules of fluid from the bright bands are drawn in amongst them; the rods are pushed farther apart; the fibre broadens with a corresponding diminution in length. This brings muscular contraction into the category of the phenomena which play the most important rôle in bringing about the varied activities of the animal mechanism. Contraction is due to osmosis. The separation of muscle into fibrils after hardening does not seem to bear out either the rod or the pore hypothesis of the structure of the dim disc. It must be remembered, however, that before the fibrils are teased apart the substance of the fibre has been coagulated. The fluid in the bright disc may thus have become as much a part of the fibril as the rod in the dim disc. The longitudinal striation of plain muscle and the appearance of continuous fibrillation in heart-muscle is more difficult to reconcile with the hypothesis that striped muscle is composed of interrupted rods.