228 PROFESSOR TAIT ON IMPACT. 



apparatus was rendered exceedingly light, strong, and compact. The lines traced could 

 easily be made as fine as those of an etching, but it was found that a slightly blunted 

 point (giving a line of about 0'005 inch in breadth) produced probably less friction, at all 

 events less irregularity, than did a very sharp one. The difference of either value of B from 

 the mean rarely amounted to more than 1"5 or 2 per cent, of the mean. When the ink 

 was dry, which happened after about a day, photographic prints were taken by using 

 the disc as a negative. [In the later experiments it was found that, when proper pre- 

 cautions were taken, no delay on this account was necessary.] To test whether the paper 

 of the positives had been distorted, in drying after fixing, a number of circles were 

 described on the glass disc at various places before the ink was dry. They were found 

 to remain almost exactly circular on the dried photograph. All the subsequent measure- 

 ments were made on these photographs. In a subsequent paper I hope to give the results 

 of careful micrometric measures, made on the glass plate itself, of the form of the trace 

 during impact. This may lead to information which could not be derived from the 

 photographs themselves with any degree of accuracy. My first object was to obtain a 

 number of separate experiments, so as to get the general laws of the phenomena, and 

 for this purpose the glass plate had to be cleaned and prepared for a new series of 

 experiments as rapidly as possible. The micrometric measures cannot be effected in a 

 short time. 



2. In the earlier experiments the fly-wheel continued in connection with the gas- 

 engine until the fall was completed. Hence the rate of rotation was irregular, and the 

 mode adopted for its measurement gave an average value only. In the later experiments 

 an electrically-controlled tuning-fork, furnished with a short bristle, made its record on 

 the disc, simultaneously with the fall of the block ; and the gas-engine belt was thrown 

 on an idle pully immediately before the experiment commenced. The angular velocity 

 of the disc was sensibly different in different experiments, according as the engine was 

 thrown off just before, or just after, an explosion. But the fact that its fly-wheel is a 

 gigantic one made these differences of small importance. They were, however, always 

 taken account of in the reductions. The disc, when left to itself, suffered no measurable 

 diminution of angular velocity during a single turn. In the earlier experiments one 

 rotation of the disc occupied about s "3 ; but I was afraid to employ so great a speed with 

 the glass plate, so its period was made not very different from one second. I found it 

 easy to obtain on the glass disc the records of four successive falls, each with its series of 

 gradually diminishiug rebounds, and along with these the corresponding serrated lines 

 for the tuning-fork. These records were kept apart from one another by altering the 

 position of the fork, as well as that of the needle-point on the block, immediately after 

 each fall. The latter adjustment alters, of course, nothing but the radius of the datum- 

 circle, and the corresponding values of the quantity A. As soon as the four falls had 

 been recorded, the glass disc was dismounted, and all the necessary details of the experi- 

 ment — e.g., date, heights of fall, substance impinged on, mass of block, &c. — were written 

 (backwards) on the printers' ink, with a sharp point, and of course appeared on the 



