350 On a Remarkable Effect produced by the 



to the sounding-line and a depth of about 2800 fathoms 

 reached. When the case came up again it looked as if someone 

 had struck it in the middle with a hammer. When it 

 opened and the cloth unfolded, it seemed for a moment as if it 

 were full of snow, but a second look showed that what appeaivd 

 to be snow was nothing but finely comminuted glass. The two 

 wider tubes had collapsed, but the narrow one was intact At 

 first sight the effect produced on the copper case was puzzling, 

 but after a little study and reflection its nature became apparent . 



No account of the experiment has been published, and when 

 I was able, through the kindness of H.S.H. the Prince of Monaco, 

 to join his yacht the "Princesse Alice" for her cruise of last 

 summer, I determined to repeat it, and, if possible, to vary it. 

 Owing to press of other work, nothing was done until near 

 the end of the cruise. The brass tube (Figs. I and 2), above 

 referred to, was the case for holding a piezometer which was 

 accidentally broken. With it I repeated the experiment 

 which I had made in the "Challenger," with this difference, 

 that I used only one sealed glass tube. It was an ordinary 

 pipette of 50 c.c., sealed up at both ends close to the body. 

 It was wrapped in a piece of muslin and loosely packed with 

 cotton waste so as to occupy the middle of the brass tube. 



The length of the brass tube was 33 cm., and its diameter 

 4-13 cm. Its weight without the cover was 350 grammes. 

 Both the top and the bottom are pierced with many holes so 

 as to allow passage to the water. 



Thus charged, it descended on the sounding-line to a depth 

 of 3000 metres, and when it came up it was evident from its 

 appearance that the experiment had succeeded. As in the 

 experiment on board the "Challenger," the glass tube had been 

 converted into a snow-white powder. The external effect also 

 was confined entirely to that part of the brass tube which had 

 been occupied by the sealed glass tube. Above and below it 

 there was no disfiguration. 



The copper ball (Figs. 3 and 4), is an ordinary 5-inch 

 ball for the supply tap of a cistern. A spherical glass 

 fractionating flask, having a diameter rather less than ij inch, 

 was hermetically sealed close to the spherical body. It was 



