226 THE AMERICAN MONTHLY [October, 



By placing the pistol in the pocket, hammer upward, and marking 

 the position of the space, I found that the latter was some two inches 

 higher than the centre of the other burned area, and that its axis was 

 inclined at a different angle from that of the said area. Experiment, 

 by firing the pistol in a pocket exactly similar in size and shape, con- 

 firming this point, and I was enabled further to swear most positively 

 that the pistol was lying hammer down in the overcoat pocket at the 

 time of its discharge. 



The next point to settle was whether the pistol was fired by hand. 

 This was easily demonstrated to have been impossible, by experiment 

 with the weapon in similar pockets, and I need not dwell upon the 

 matter here. 



How far, in all probability, was the muzzle of the pistol from the 

 point of exit of the ball, was the next vital question. This was also 

 solved experimentally ; but before the experiments could be accepted as 

 evidence the microscope had again to be resorted to. The silk lining 

 of the coat had not been attached to the hem of the garment, and hence 

 it was a comparatively easy matter to turn the latter wrong side out. 

 On the one side was the silk and cotton stuff' used for coat lining, on 

 the under side of which cotton predominated. A strip of buckram ex- 

 tended some six inches from the margin of the coat inward and pro- 

 tected the satin trimming from the action of the flame. The buckram, 

 from its closely twisted nature, showed but a slight discoloration, which 

 might or might not be due to flame. The lining also showed but slight 

 traces of discoloration. Coming to the chinchilla, however, there was 

 an area of about six inches in diameter that, while showing no char to 

 the unaided eye, had evidently been subjected to some denuding influ- 

 ence. 



Placing the surface as flat as possible and in a strong direct light, I 

 first went over it with a Coddington lens, but beyond the fact that there 

 were points where the threads were cut and broken by the impact of 

 some substance, I arrived at no direct and positive evidence which 

 would enable me to sv\^ear positively that the denudation of this area 

 was due to flame. The round bullet hole in the centre and the charac- 

 ter of the other marks rendered this conclusion very probable, it is true, 

 but absolute certainty was what I wished to obtain. 



Again, before resorting to the expedient of secret mutilation of the 

 garment, though never so slightly, I resorted to experiment. Taking 

 a piece of chinchilla exactly similar to that used in the overcoat, prob- 

 ably a remnant of the same bolt, as it was procured from the tailor 

 who made the garment, I hung it up in such a manner that it would 

 ofl^er the same resistance to the passage of a bullet as the stuff' in the 

 garment. I then put the pistol in a pocket exactly similar to that in 

 the coat, and discharged it by a slight blow upon the hammer (which 

 latter rested, of course, upon the cartridge, which latter was one ex- 

 actly similar to that in the pistol at the time of the accident). The 

 muzzle was about six inches from the chinchilla. The discharge made 

 the characteristic marks in the pocket, the characteristic tear in woollen 

 and cotton-twill material, bored a ragged hole about three-quarters of 

 an inch in diameter through the chinchilla, and made a burn about 

 four inches in diameter. Taking the chinchilla and rubbing it until 

 pretty much all the char had been removed, I passed it backward and 



