291 Prof. R. E. Salmi on the 



on the developed plate was a small and dense spot, with 

 increasing pressures it became larger and at the same time 

 less and less dense. The impression at very low pressures 

 was almost uniformly dark except at the extreme edges, but 

 with higher pressures there was a gradual falling off in the 

 density of the impression from the centre outwards. With 

 different gases the size of the impression under similar con- 

 ditions was different. Plate VI. A, B show microphotographs 

 of two impressions obtained with air: A under a few mm. 

 pressure, and B under 161 mm. The two impressions were 

 taken under conditions which were otherwise similar, and 

 both have been enlarged to the same extent. They bring out 

 in a fairly striking manner the differences just noted, 



Method of counting the Silver Grains. 



For counting the grains, a microscope with ^ in. objective 

 and x 8 eyepiece were used. The eyepiece was provided 

 with a micrometer in squares, so divided that each square 

 covered exactly 400 sq. /j, of the magnified object. A thin 

 sheet of mica with two very fine cross-lines scratched on it 

 was placed on the photographic plate with the crossing of 

 the lines exactly over the centre of the image, this adjustment 

 being first made roughly with the naked eye, and afterwards 

 more accurately with the help of the microscope. The 

 number of grains in each square of the eyepiece micrometer 

 along one line from the centre outwards was then noted, the 

 plate being shifted along that line as the periphery of the 

 image was approached. The scratched lines over the mica 

 are too far out of focus to interfere with the counting, while, 

 at the same time, serving as reliable guides. In order to 

 protect the photographic film from dust and scratches, it is 

 important to cover the impression with a drop of Canada 

 balsam and a thin glass coverslip as soon as the film is dry. 

 The want of this simple precaution cost the writer the loss of 

 several valuable plates. The use of an oil-immersion lens 

 was found undesirable for several reasons. Besides greatly in- 

 creasing the labour of the counting, it actually multiplies the 

 chances of error in the process. For, whereas in the case of 

 the ^ in. objective the grains at different levels in the film are 

 nearly all visible at one focus, with a T ] g in. objective they 

 require different adjustments, and very often it becomes 

 difficult to say whether a particular grain has been already 

 recorded, unless all the grains are first drawn by means of a 

 camera lucida and then counted. 



The centre of the impression having been adjusted carefully 

 in the manner already explained, the number of grains seen 

 within successive equal area's of 20 x 20 /jl are counted from 



