168 Mr. R. Threlfall on the 



gauges and the wooden bar supporting the charge. The 

 gauges were always equidistant from the explosion centre. 



A key placed in a convenient position enabled me to fire 

 the charge whenever I desired, without distracting my atten- 

 tion from the reading of one of the gauges. The other gauge 

 was read by Mr. Walker, who soon learned the way of it and 

 became an accurate observer. The results were at first sur- 

 prising. It will be noticed that extreme precautions were 

 taken to make the distribution of effect about the vertical axis 

 through the point of ignition as symmetrical as possible. 

 In spite of these precautions I found that the gauges by no 

 means kept pace with each other. First one would give a 

 large reading, and then the other. The average reading was 

 about 3 inches of coloured water. I finally traced this out 

 to the almost unavoidable unevenness in the glass of the bulbs. 

 A new set of bulbs were therefore blown of thinner glass, and 

 carefully examined as to uniformity before being used. The 

 readings now became more consistent, though not so much so 

 as I should like to have seen them. While considering how 

 to make certain of a more perfect symmetry, I discovered a 

 peculiar action which finally caused me to abandon this form 

 of gauge entirely, as I could not be sure about its theory. I 

 had imagined that the energy of explosion would be trans- 

 mitted by the almost unstrained layer of thin india-rubber 

 without appreciable loss to the interior of the gauge, and 

 that the result of the impact would be a rise of the liquid in 

 the tube, much as if the gauge had been struck on its face 

 by a hammer. Now the tambourine -shaped end of the gauge 

 and about 6 inches of the connecting-tube projected freely 

 into the water below the wooden support, so that the dis- 

 tance from the back of the gauge to the side of the tank 

 was about 4 inches. 



After several shots had been fired, I noticed that both the 

 tubes of the gauges were bent towards the centre of the ex- 

 plosion, not away from it. A wedge placed behind the tam- 

 bourine and between it and the side of the tank, and pressed 

 tight in, increased this effect. How it came about, I was at 

 a loss to discover ; but the result was that I decided I knew 

 too little about the motion of the water to trust to the proper 

 deformation of the india-rubber. The rapidity with which the 

 coloured liquid rose and fell was also a drawback. The gauges 

 were troublesome to set besides, so that I finally determined 

 to make two new ones. In these I adopted a pendulum method. 

 The two ends of a bit of brass tube, one inch in diameter 

 and four inches long, were closed by caps. Through the 

 centre of these caps I bored holes three eighths of an inch in 

 diameter and immediately opposite one another. The tube 



