88 REPORTS ON THE STATE OF SCIENCE. 
chamber twenty yards from the mouth of a tunnel driven into the chalk, 
a hole about two feet square was excavated. Into this a box with a light- 
proof door was cemented. In this box a Richard self-recording ther- 
mometer drum covered with a Kodak bromide paper was placed. The 
drum, which was entirely made of brass, revolved at the rate of 40 milli- 
metres per day. Between it and the chalk face, distant about half an 
inch, there was a sheet of zinc in which were two round holes respectively 
about an eighth and a quarter inch in diameter. One was below the 
other, and lower still there was a vertical slit. As the drum turned, the 
paper was exposed to the chalk through these openings. Experiments 
were continued for four months, ending May 8, 1903. During that interval 
photographic impressions were only obtained on three or four occasions. 
These were in the form of black discs, possibly representing the holes and 
straight lines for the slit. The first series read as follows: February 6, 
6.30 p.m., 8.54 p.m.; February 7, 0.30 a.m. and 2.54 a.m. Nothing in 
the nature of a glow extending over several hours or anything coinciding 
in time with a large earthquake was recorded. It was, however, interest- 
ing to note that photographic effects had been obtained in a place and 
under conditions where it is difficult to imagine that they had been the 
result of artificially produced light. 
For various reasons these experiments were not again taken up until 
August 19, 1906. On that date a piece of apparatus in many respects 
similar to that used in 1903 was put up in a dark chamber, cut in the 
chalk inside the tunnel leading to the Pan Chalk Pit. The chief differ- 
ences between the new and the old installations were as follows: In the 
new apparatus the cylinder carrying the paper moved at the rate of 90 mm. 
per day. This shortened the exposure of the film as it passed before the 
holes in the zine plate between it and the chalk face. The distance 
between the film and this plate was reduced to one-eighth inch, whilst an 
aluminium rim forming the bottom of the brass drum revolved inside a 
horizontal slit cut in the plate. The rim was at a distance not greater 
than one sixteenth inch from the top and bottom of the slit. 
The drum stood inside a wooden case outside which at the distance of an 
inch was a second case. The dimensions and form of the holes in the 
zinc sheet which formed the end of the inner box were as follows: A 
round hole one-eighth inch in diameter, a triangular hole with half-inch 
sides vertically below the round one, and below this a square hole with 
sides of one inch. The holes were about an inch apart, and underneath 
the square hole was the slit, cut to free the rim of the drum. 
The movement of a small electric lamp round the face and sides of the 
box produced no effect on the paper inside, and hence it may be inferred 
that the box was light-proof. A self-recording thermometer and a hygro- 
meter between November 5 and 19 showed that the temperature and the 
moisture in the chamber were practically constant. 
A similar piece of apparatus was, through the kindness of Mr. B. 
Angwin and Mr. J. G. Lawn, placed in the King Edward Mine, Cam- 
borne, Cornwall, at a depth of 160 feet. The rock was damp, as in Pan 
Chalk Pit. At both places cakes of calcium chloride were used to dry the 
atmosphere, but I cannot say they were effective. For about three months 
papers were exposed simultaneously with those at Shide, and the results 
were compared. Another piece of apparatus was set facing the chalk in 
a light-proof hut in White Pit Lane, between Shide and Carisbrooke. The 
drum was inclosed in a box (also light-proof) inside the hut. Its rate of 
