710 



SCIENCE. 



IVoi.. II., No. 43. 



seen at the bottom of the picture. We see that 

 the interval is not always the same, and' that, if 

 ■we suppose equal intervals of time to separate 

 successive images, the greatest velocity occurs in 

 the run which preceded the leap, and that there 

 was a diminution of speed while the leaper was in 



the air: in fact, this diminution is still increased 

 after the fall, the velocity being partly lost the very 

 moment the body touches the ground. In order to 

 know whether the images have been produced at 

 equal intervals of time, and the duration of these 

 intervals, the dial of the chronograph must be con- 

 sulted. It is then seen that the luminous needle is 

 represented as many times as there have been illu- 

 minations, namely, nine times; that the interval be- 

 tween the illuminations was uniform, for the images 

 of the needle whose rotation was 

 uniform form equal angles : in short, 

 the absolute value of the time-inter- 

 vals between the illuminations is 

 expressed by the angle which the 

 images of the needle on the dial 

 make. This angle is about 86°, 

 which shows that the time-interval 

 between successive illuminations is 

 one-tenth of a second. From these 

 measures of time and space we easily 

 deduce the velocity of the leaper in 

 the various phases of the experiment. 

 This speed was seven metres a sec- 

 ond during the preliminary run, five 

 metres during the leap, and fell to 

 three metres and a half after the fall. 

 When one takes on the same plate 

 a series of photographs representing 

 the successive attitudes of an animal, 

 he naturally tries to multiply these 

 images in order to know the largest mmiber possible 

 of the phases of movement ; but when the latter 

 is not rapid, the frequency of the images is soon 

 limited by their superposition, and by the confu- 

 sion which results. Thus, a man running even at 

 moderate speed can be photographed nine or ten 



times a second (fig. 0) without confusion of the images. 

 If sometimes one limb is depicted where the other limb 

 has already left an impression, this superposition does 

 not at all affect the images: the wliite only becomes 

 more intense where the plate has received two im- 

 pressions, so that the contours of the two members 

 are still easily distinguishable. But 

 when a man walks slowly, as in fig. 

 7, the images present so many su- 

 perpositions that confusion results. 

 This inconvenience is remedied by 

 partial photography ; that is, by sup- 

 pressing certain parts of the image 

 in order that the rest may be more 

 easily distinguished. As by our 

 method white and bright objects 

 only make an imiiression on the sen- 

 sitive plate, it is necessary merely to 

 clothe in black the parts of the body 

 which we wish to exclude from the 

 image. If a man dressed half in 

 white and half in black is walking 

 on the track, and turns toward the 

 photographic apparatus the part 

 clothed in white, the right, for in- 

 stance, there will appear in the pic- 

 ture only the right half of his body. These images 

 allow us to observe in their successive phases, first, 

 the pivot-like turning of the lower limb around the 

 foot during the time of support; and second, as 

 the foot rises, the turning of the same limb around the 

 hip-joint, while at the same time this joint is moving 

 forward without cessation. 



Partial photographs are also serviceable in an anal- 

 ysis of rapid movements, because by this means the 

 number of attitudes represented may be many times 



Fig. ' 



increased. Nevertheless, when the image of a limb 

 is moderately large, the partial photographs cannot 

 be too much increased without confusion through 

 superposition. We must therefore diminish the size 

 of the image, if we desire to repeat them at very short 

 iiitervals. For this purpose thf walker is clothed 



