OF ARTS AND SCIENCES. 479 



Several observations were made in each case, and it was found that 

 they always agreed with one another within less than ten per cent. 

 Tlie reason why the observed exposures ai*e somewhat shoi'ter (about 

 twenty per cent, in this case) than the theoretical ones is, that even a 

 very brilliant object does not begin to produce a photographic impres- 

 sion upon the plate until quite a large portion of the lens is uncovered. 

 The exposures are therefore somewhat shortened. Tliis effect, which 

 would be added to that of the friction introduced by the style, would 

 seriously interfere with the accuracy of the method published a few 

 months ago, of allowing a tuning-fork to trace a sinusoid on the 

 smoked surface of the drop itself. 



It was suggested in the British Journal of Photography, as early as 

 August, 1883, to photograph a freely falling glass ball placed in the 

 sunlight. Applying to this the laws of falling bodies, the exposures may 

 at once be calculated. The suggestion occurs at once that the resist- 

 ance of the air might be of suthcient importance to vitiate entirely tlie 

 results thus obtained. In order to ascertain whether this was the case, 

 the following experiment was devised. A glass ball, silvered on the 

 inside, such as is used for Christmas trees, was procured. It measured 

 4.15 cm. in diameter, and weighed 25 grains. Some white silk threads 

 were attached to a blackened board so as to form a scale of equ;il parts, 

 and, this being set in the sunlight, the ball was dropped by its side into 

 a box filled with cotton wool. A mirror was attached to the side of a 

 tuning-fork, and the pitch determined. A camera was now placed so as 

 to photograph the image of the ball as seen in the mirror. The fork 

 was set in vibration and the ball dropped. On development, the plate 

 showed a black sinusoidal line, the vibrations of which were near 

 together at the top, but gradually widened out as the ball approached 

 the bottom of its course. Several photographs were taken, and meas- 

 ured under a dividing-engine, with the following results. 



As it was impossible to determine the exact time of starting, the 

 nearest point which could be precisely measured was selected. Its 

 distance from the starting-point was called s, , and the distance of 

 some other point somewhat lower down was called s^. The distances 

 that would have been traversed if the air had offered no resistance are 

 designated by S^^ and S.^. The number of vibrations executed by the 



