156 



SCIENCE. 



[Vol. II., No. 27. 



Two foiins of apparatus for collecting sam- 

 ples of water from different depths are shown, 

 both constructed on the same principle. The 

 larger, having an ingenious means of closing, 

 is chosen for description here. It consists 

 of a brass cylinder open at both ends, about 

 ten inches in length by four and a half in diam- 

 eter, sliding freely through a space somewhat 

 greater than its length, between three vertical 

 brass rods or guides, which also constitute the 

 frame of the apparatus. AVhen the cylinder 

 slides down, it encloses a vertical rod having 

 a horizontal plate at the top, which forms a 

 tight cover for the cylinder, similar to the end 

 of a piston. The bottom of the cylinder falls 

 into an annular groove in which a sheet-rubber 

 ring is fitted, thus making a tight joint below. 

 A rubber ring is also employed to make the 

 upper joint tight. In the smaller instrument 

 the lower groove is filled with a mixture of 

 suet and wax ; and the cylinder has an annular 

 plate on top, the border of which extends in- 

 wards sufficiently to be bent downward so as 

 to fit into a similar groove on the upper surface 

 of the horizontal disk forming the top of the 

 closed chamber. When the apparatus is sent 

 down, the cylinder is suspended at the top. 

 When it reaches the desired depth, the cylinder 

 is released by a mechanism to be described, 

 and falls, enclosing a sample of the water. In 

 the smaller apparatus the cylinder is sustained 

 during descent by the resistance offered b}- the 

 annular plate above referred to, which is con- 

 siderably larger than the diameter of the cyl- 

 inder. On drawing up the apparatus, the plate 

 also acts to force the cylinder well down into 

 the grooves. In the larger instrument the cji- 

 inder is held up by a catch, actuated by a 

 system of levers, which are connected with 

 a turbine wheel enclosed in a brass case at the 

 top. During descent the water passes through 

 the case, entering and leaving it freely through 

 strainers of brass gauze, and causes the tur- 

 bine to revolve. The latter turns freelj' until 

 the desired depth is reached. When ascend- 

 ing, the wheel makes a certain number of revo- 

 lutions in the op()osite direction, and soon acts 

 upon the system of levers through a ratchet 

 and ratchet-wheel, thus releasing the cylinder. 

 This instrument has been successfuUj- used in 

 depths of three hundred metres. It is suffl- 

 cientlj' good to enable the quantity. of air con- 

 tained in the water at different depths to be 

 determined. 



Arfwidson's water-bottLe, exhibited by Den- 

 mark, is a simple cj'linder of brass, shaped 

 somewhat like a bell, closed by bottom and top 

 plates with bevelled edges ponnepted by a cen- 



tral stem. The bell falls,' and the whole appa- 

 ratus is drawn up by the central roil. The 

 joints are made tight by grinding the plates 

 and cylinder together. It is very simple, very- 

 light, and seems to be a good instrument. No 

 information concerning its use is available at 

 the present moment. 



Another of Professor Ekman's instruments 

 is used to collect samples of water, and also to 

 enable the temperature to be cori'ectly deter- 

 mined. Although quite different in construc- 

 tion from the others, it is the same in principle, 

 except that it is made to protect the sample 

 from any change of temperature while being 

 drawn up, so that a thermometer may be in- 

 troduced on deck to get the temperature of 

 the stratum of water from which it was taken. 

 The instrument has been found to give accu- 

 rate results at depths of two hundred metres. 

 It is not stated whether it has been used at 

 greater depths. 



In this instrument the cylinder is fixed be- 

 tween two galvanized iron rods, which, with 

 four horizontal circular bands of the same ma- 

 terial outside, constitute the frame, resembling 

 a sort of cage. The top and bottom of the cyl- 

 inder are formed by what may be described as 

 two piston-heads connected together by a hol- 

 low rod, which slides up and down on anotlior 

 rod running vertically through the middle of 

 the apparatus. The piston-lieads are made 

 of thick gutta-percha secured between brass 

 Ijlates. The connecting-rod is also covered 

 with gutta-percha, and the cylinder itself is 

 lined with it. Rubber is used to make the 

 joints perfectly tight. The sample of water is 

 thus protected by gutta-percha in every direction 

 about two and a half centimetres in thickness. 

 The upper piston-head carries a brass plate, 

 which offers sufficient resistance to the water, 

 while descending, to sustain it at the top of 

 the apparatus. On hauling in, the water 

 forces the piston down into the cylinder, en- 

 closing the sample. The apparatus gives re- 

 markably good results, if we may judge from 

 some of the figures given in the case of a 

 series of temperatures taken in the Baltic, 

 where the alternations of cold and warm strata 

 were quite remarkable. The temperatures 

 were recorded to tenths of a degree of Celsius's 

 scale, as, indeed, it was necessary that they 

 should be, in order to make the results of un- 

 questionable value ; for the total variation in 

 temperature between depths of 50 metres 

 (when the temperature was 1°.8) and the 

 bottom, 210 metres, was only 2°.l C, yet 

 there was a rise to 3°. 9 at 100 metres, and a 

 fall to 3°.l at 210 metres. 



