544 
Proceedings of the Royal Society 
it were ground to fit on the conical edges of the base plate and the 
upper plate. Strong metal guides fixed to the shank supporting 
the base plate ensured that the cylinder should slip fairly on its 
fittings. On striking the bottom the tumbler threw off the slip, 
and the bottle closed. The “Challenger” instrument worked in 
the same way. To slip the cylinder at intermediate depths the 
“ Challenger ” investigators used a plate resembling that employed 
for closing the stop-cock bottle, and Meyer adopted a device better 
adapted for working in shallow water. He suspended the slip- 
cylinder by a cord and two hooks to two horizontal pins fixed to 
the upper part of the apparatus. A split wedge attached by rods to 
a ring surrounding the sounding-line rode on each pin; when a 
weight was let down the line from the ship it depressed the ring, 
which lowered the wedges and pushed the hooks off the pins, so 
closing the apparatus. 
The superiority of slip water-bottles lies in their being adapted to 
take water from a perfectly definite position. In fact, a slip water- 
bottle, when it acts properly, cuts out of the layer in which it closes 
a cylinder of water, the parts of which retain their relative positions 
until the water-bottle is emptied. 
In Buchanan’s small slip water-bottle* the cylinder is allowed 
to run down the line from the ship. Its defects are its aptness to 
indent the edge by striking on the brass hemisphere, and its being 
too light to press sufficiently on the rubber to ensure that its 
contents are unmixed. It cannot, of course, be used on the same 
line with thermometers. This instrument was at first used in the 
work carried on by the Scottish Marine Station on the Firth of 
Forth, where water samples have to be obtained at all depths from 
2 fathoms to 40, sometimes in a strong current, sometimes in a 
rather rough sea, and sometimes in a narrow channel, where there 
is considerable traffic, and where the instruments cannot be left 
down for many minutes. It was not very well suited for this work ; 
for which it was found desirable to have a water-bottle fulfilling 
the following conditions : — 
1. To be light enough to work by hand when necessary. 
2. To contain about 2 litres of water. 
3. To be used on the same line with thermometers. 
* Described and figured, Proc. Roy. Soc. JSdin., xiii. (1885), p. 32. 
