AN INVESTIGATION INTO THE BREAKING OF WATCH 
MAINSPRINGS IN GREATER NUMBERS IN THE 
WARM MONTHS OF THE YEAR THAN IN THE 
COLD MONTHS. 
RICHARD G. NORTON. 
(With Plate XXXV). 
In looking over my watch-repairing 1 record, I find that for 
a period of ten years, from January 1, 1880, to January 1, 
1890, nearly twice as many mainsprings were broken during 
the months of April, May, June, July, August and September 
as were broken during the months of January, February, 
March, October, November and December. For another pe¬ 
riod of ten years, from June 10, 1893, to June 10, 1903, the 
record shows a total of i693 springs broken, of which 449 broke 
during the warm months of April, May, June, July, August 
and September, and 244 broke during the cold months. 
It is supposed by many persons that mainsprings break in 
greater numbers during electrical storms than at other times; 
but I have good reason to think that electricity has little or no 
influence in the matter, since clock springs would be expected 
to break from the same cause. There are as many clock springs 
in use as watch springs, and probably more, and I am sure 
that during my practice of repairing watches and clocks not 
more than a dozen clock springs have broken to one hundred 
watch springs. In fact, there have been.periods of a year dur¬ 
ing which I have not had occasion to replace a broken clock 
spring. 
It is a well-known fact that platers of watch cases remove the 
steel springs from the cases before placing the cases in potas- 
