1881.J 407 [Chance. 



Months. „ ^°- ¥ 



Explosions. 



June 10 



July 4 



August 3 



September 2 



October 4 



November 3 



December 8 



Total 45 



This list embraces only the explosions resulting in great loss of life. The 

 minimum loss of life was 20, and the maximum, the Oaks Mine disaster, 

 December, 1866, was 362 ; the loss of life aggregates 2621, an average of 

 more than 58 for each disaster. 



The occurrence of three of these explosions on June 2, 1862 fat Wash- 

 ington, Guindrseth and Coppal), and two of the most fatal on December 

 12th and 13th, 1866 (the "Oaks" and "Talk o' the Hill"), by which 362 

 and 92 lives were lost, certainly point to atmospheric disturbance as the 

 immediate cause. The occurrence of a large percentage of these disasters 

 at semi-annual periods, June and December, seems to indicate the occur- 

 rence (in Great Britain) during these months of unusually high barometer, 

 followed by a decided fall, as the probable cause of these great outbursts 

 of gas. 



But the problem I have been considering is somewhat different, for the 

 table embraces all the explosions, whether large or small, occurring during 

 the ten years. Is shows a decidedly larger number for the warm than for 

 the cold months, and therefore points primarily rather to impairment of 

 ventilation from high temperature than to barometric changes as the true 

 cause of the diflference ; but the occurrence of two maximum periods. May 

 and October, seems to indicate that barometric changes have also exei'cised 

 an important influence on the relative efflux of gas. 



The amount of rise and fall does not seem to have a perceptible effect, 

 for the monthly barometric runge is greatest during the cold mo7it7is, whereas 

 fire-damp explosions are most frequent during the warm months. Fre- 

 quent and abrupt changes from high to relatively low barometric pres- 

 sure, are the probable cause of many explosions, though the movement of 

 the mercury may not amount to more than one-eighth or one-quarter of 

 an inch. An unusually high barometric column is always an intimation of 

 coming danger. 



