542 
Proceedings of the Royal Society 
In very shallow water a bottle provided with a glass stopper may 
he sunk, and the stopper pulled out by a second line when at the 
proper depth. Stevenson’s hydrophore (1813)* is an improvement 
on this. It is a metallic bottle closed by a plug, which may be 
removed and replaced at any depth, in order to fill the vessel and 
to secure the sample unchanged. Stevenson’s deep-sea hydrophore f 
resembles one of Marcet’s forms. It is a spindle-shaped vessel, 
fitted with two conical valves on a very heavy axis, which projects 
beneath the apparatus and keeps it shut during descent by its 
weight. When it strikes the bottom the valves are forced up, the 
air escapes, and water enters. On raising, the valves fall back into 
their places, and do not tend to open. 
Jacobsen J employed, during the cruise of the “ Pommerania,” in 
1872, a glass cylinder adequately protected and provided with 
weighted valves which hang open during descent ; at the required 
depth it is inverted by a second line, and the weights pressing on 
the valves serve to secure them. The difficulty of employing a 
double line is so great, when the sea is rough or when currents are 
rapid, that apparatus requiring such fitting is very inconvenient. 
The Commission at Kiel, for the physical, chemical, and bio- 
logical examination of the German seas, has devised many methods 
of collecting water.§ When the analysis of the dissolved gases in 
the sample was contemplated, an india-rubber bag containing a little 
mercury and squeezed free from air, was sunk and opened at the 
proper position. 
On the French “ Travailleur ” expedition of 1881, a water- 
bottle combining the peculiarities of the stop-cock and valved 
forms was employed.|| On the “ Talisman ” expeditions a strong 
glass globe was exhausted by an air-pump, and its capillary 
opening sealed ; it was fixed to the line below the Negretti-and- 
Zambra thermometer, which when reversed broke the tip of the 
capillary tube and water entered the globe. When the globe was 
received on deck the water sample was permanently secured by 
sealing off the broken capillary. No other process of collecting 
and preserving samples of water is so perfect theoretically as this, 
* Described and figured, David Stevenson’s Canal and River Engineering 
(1872), p. 125. t Described and figured ibid., p. 128. 
X Liebig’s Annalen y *c\xvii. (1873), p. 1. 
§ See their annual Berichte, passim . [| La Nature , 1882, i. 53. 
