1890. J NKW YORK ACADEMY OF SCIENCES. 101 



pictures at first obtained only approximately afforded the 

 expected result. The flame at the mouths of the cases pro- 

 duced a marked effect, but the trajectories of the sparks were 

 less distinctly traced, owing either to their smallness, rapid 

 flight, or low incandescence. Four-ounce triangles, rosette 

 wheels, China flyers, and ten-inch vertical wheels all gave outer 

 fringes in which the detail of the sparks was easily obliterated 

 by long exposures, owing to the number and superposition of 

 their images. 



Seeking to obtain small wheels affording large and brilliant 

 sparks, like those emitted by exhibition pieces, I found that the 

 sparks from these were due to steel filings, which are not put in 

 commercial wheels, as they are liable to be destroyed by rust 

 before such pieces are used. From ten-inch vertical wheels 

 containing steel filings, zinc filings, and powdered magnesium, 

 made especially for me by the Unexcelled Fireworks Company, 

 very attractive pictures were finally obtained. The so-called 

 pot usually attached to small wheels, and consisting of a case of 

 Bengal fire, was found to overpower the discharge from the 

 driving-cases both by halation and illumination of its own 

 smoke, unless it was arranged to light at the last moment and 

 was photographed by a separate short exposure. From these 

 experiments, it was concluded that the ideal small wheel for 

 photographing should emit sparks about one-tenth of the size 

 of Roman-candle balls, and might have one or two pots arranged 

 to light without an explosion and burn only during one revolu- 

 tion of the wheel. 



The ''ghost," or reflection from the lenses, which is of less 

 moment in ordinary photography, becomes so conspicuous in 

 pyrotechnical photography as to be a matter of considerable 

 solicitude. Views of wheels were presented in which an ad- 

 ditional image resembling the circular ornament in the centre 

 of the wheel appeared near the edge of the plate. That its 

 circular form is not dependent on that of the primary image, 

 was shown by pictures in which candle and lamp flames photo- 

 graphed by short exposures were accompanied by very distinct 

 circular impressions obtained by longer exposures. In making 

 some of these views, screens were used close to the plate-holder, 

 to cover portions of the plate. The effects and the avoidance 

 of these reflections in pictures by artificial lights, were discussed 

 at length and illustrated by portraits taken by lamplight. 



The next following views, showing brilliant haloes of several 

 forms, obtained from the sun with various lenses, illustrated 

 that the so-called halo which surrounds the disc of the sun 

 when it is located in or near the centre of the plate, moves from 

 the centre of the plate in one direction in proportion as the sun 



