118 ANIMAL MECHANISM. 



sontcd by llio line O P0 (fig. 22). It is seen that the pubis 

 rises at the middle of the pressuic exerted by each foot, and 

 sinlcs at the instant when the weight of the body passes from 

 one foot to the other. 



The real amplitude of those oscillations is about 14 milli- 

 metres, according to MOJIS. Carle t. This movement, however, 

 varies with the length of the step; it increases with it, but 

 this increase docs not depend on the maxima of the curve 

 being more elevated, luit on its minima being lower. 



AVe may explain these phenomena very easily. When the 

 body is about to quit the support of one leg, this limb is in 

 an inclined position, and the result of its obliquity is that its 

 superior extremity which sustains the trunk is at a less height. 

 The other leg, which reaches the ground at this instant, is 

 slightly bent ; it will soon draw itself up, and thus raise the 

 body which is supported by it; but in this movement, the 

 leg describes the arc of a circle around the foot resting on the 

 ground; therefore, in the series of successive positions which 

 it occupies, the body rises more and more as the leg which 

 supports it approaches the vertical position; it sinks again as 

 the ley becomes oblique. 



We can easily peiceive that the length of the step lowers 

 the trunk, by increasing the obliquity of the legs. Indeed, 

 the constant character found in the maxima of the vertical 

 oscillations is explained by this fact, that the leg, when ex- 

 tended and vertical, constilules necessarily a constant height 

 that, which answers to the maximum of the elevation of the 

 body. 



Horizontal oscilltilioni of tlte body. The pubis, since that is 

 the point whose displacement we are now studying, is carried 

 alternately Irom left to right, and from right to left, at the 

 same time as it moves vertically. In order to register these 

 movements, we make use of a lever-drum arranged in such a 

 manner that the membrane is forced inwards and outwards 

 alternately by the lateral movements which are given to the 

 lever. During this time the registering lever, connected with 

 it by means nf tin; tube, oscillates vertically, in which direc- 

 tion alone tracings e,m he made on the cylinder. If, in the 

 curve which is traced, the elevation corresponds with a trans- 



