ja2 REDORT—1896. 
APPENDIX I. 
Eniarged Shadow Photographs of Screws. By Col. Watxty, C.2., R.A., de. 
The objects aimed at in producing enlarged photographs of screws 
are— 
(a) To provide a means of verifying the accuracy of the shape of a 
thread. 
(6) To provide a means of accurately gauging the dimensions of screw 
threads. 
(c) To provide a “record or certificate of a screw, in a similar manner 
to the certificate of accuracy given by Kew Observatory. 
As in this process a standard scale is photographed at the same time 
as the screw, a direct reading to any desired accuracy can be obtained. 
There seems to be hardly any limit to the amount of enlargement, as 
the difficulties inherent to the enlargement of an ordinary negative do not 
apply to this process. 
The question arises, Does the shadow photograph give accuracy as 
regards— 
1. Dimensions ; 
2. Shape of thread ? 
To test (1) a No. 2 B.A. thread was enlarged 37:18 times. The linear 
dimensions could in this photo be measured to at least soooth of an inch. 
The diameter, pitch, and angle of thread of this particular screw were 
gauged by a member of the Committee, and found to be: 
By an Elliott gauge . - - 01818 alten (os 
» &@ Brown and Sharp gauge , 071820 Sa 
0:03184 pitch 
54° 30’ angle of thread 
The photo gave the following :— 
01818 diameter 
0:03181 pitch 
54° 30’ angle of thread 
Fig. 1 shows a reduced copy of an actual photograph and scale. 
As regards (2) there has been considerable discussion as to whether 
the shadow photograph gave the true shape of the thread. There are 
two methods of obtaining a 
photograph of a screw thread = 
one in which the axis of the 
screw is at right angles to the 
beam of light ; the other im 
sp which the axis of the screw 
is inclined at an angle which 
differs from the right angle 
by the angle of the pitch of 
the screw. 
D Asregardsthe first method, 
a mathematical consideration 
worked out by Mr. Price seems to show that a slight correction of about 
4 degree in the total angle of the screw thread would have to be made. 
i think, as regards the second method, no correction is required. In any 
