354 O n a Remarkable Effect produced by the 



not occupy a truly axial position, but lay nearer that part 

 of the brass envelope where the ears for attachment to the 

 sounding-line are situated. Here a most formidable corrugation 

 (Fig. i) has been produced, the metal being pinched into a fold 

 so as almost to meet inside. Besides this, there are two minor 

 corrugations. A greater thickness of water intervened between 

 this part of the brass envelope and the enclosed glass tube, and 

 the small effect produced shows that the difference of pressure 

 within and without the brass tube was here comparatively 

 small. It will be observed that the butt-joint of the tube has 

 been opened at Fig. 2 ; but this is a secondary effect due to the 

 distortion. 



The brass tube, as it stands, is a manometer or pressure 

 gauge which records the distribution of pressure in it while 

 filled with and immersed in water, during the instant of time 

 when, while the pressure on all sides is very great, the pressure 

 at a locality in the interior suddenly becomes nothing or very 

 small. The effect of this sudden difference of pressure has 

 been concentrated on the part of the brass tube nearest to 

 which the glass tube was situated. Here the diminution of 

 internal volume of the brass tube produced by the principal 

 corrugation must, from rough measurements, be very nearly 

 equal to that of the glass tube which collapsed. At first sight 

 it appears remarkable that on the collapse of the glass tube, 

 when it was free to the compressed sea-water to fill up the void 

 with water through the two open ends, instead of doing so, it 

 filled it by pinching up the stout brass of which the tube 

 made, to such an extent as to obliterate the void. 



The experiment shows us that it was easier in the time t<> 

 pinch the envelope of brass than to shove in the plugs of water 

 at both ends. The complete absence of distortion or disfigure- 

 ment of the upper and lower portions of the brass tube shows 

 that the tension of the water in these two portions of the tube 

 was not materially diminished in the time between the collapse 

 of the glass tube and the occupation of its place by the corru- 

 gation of the envelope. In considering this experiment, we 

 must distinguish between the tension and the pressure of the 

 water. When the water is at rest they are equal. During a 



