378 



NATURE 



\A7igtist 19, 1880 



But this pressure produces a rise of 4361 divisions on manometer 

 No. 2. We have then for the value of one division on the 

 manometer — 



. = '36^6 ^ 

 4361 

 Hence to convert readings of manometer No. 2 into atmospheres 

 we have to multiply by 3-132 the difference between the mano- 

 meter reading under pressure and that at atmospheric pressure. 



In another series of experiments piezometer K, No. 4, was 

 compared with manometer No. 2, both being at a temperature 

 of 12-5° C, and the following results were obtained as the mean 

 of nineteen observations : — 



Mean rise of manometer No. 2... (A) 41-35 divisions. 

 Mean apparent contraction per \ 



thousand of water in piezometer > (H) 5-8782 



K, No. 4 ) 



But from the results in Table I. Ave have for the pressure m 

 atmospheres — 



P = 3-132 X A = 3-132 X 4I"35 = '29"S atmospheres. 

 And the apparent compressibility of water in glass at this tem- 

 perature (12-5° C.) in volumes per thousand per atmosphere is— 

 \\ 5 '8782 ^ 



M = 



129-5 



= 0-04539. 



Dr. William Robertson, who had very carefully verified its 

 graduations. It is remarkable as a coincidence that the values 

 of the divisions turned out to be identical, namely, 0-000417". 



In the observations recorded I made no attempts to subdivide 

 the micrometer divisions further than to estimate a half. As 

 the micrometer readings are not affected directly by the pressure. 



We see then that at pressures up to 240 atmospheres the property 

 peculiar to water of diminishing in compressibiUty with rise of 

 temperature is preserved unimpaired, and the amount of change 

 corresponds closely with that found at low pressures in the 

 e.xperiments of Regnault and Grassi. . , r 1, 



In Table II. the results obtained are summarised. In the 

 first column we have the number of the series, and in the second 

 the number of observations which constitute the series. Under 

 T we have the temperature of the receiver and therefore of the 

 rod in it. The experiments were made at the temperature of 

 the room, which varied very slightly. The arithmetical mean 

 of the values of T is i2'-77 C. Under A we have the pressure 

 in terms of the scale of the manometer ; that is the difference 

 between the readings of the manometer when the pressure was 

 up and when it was equal to that of the atmosphere. ^ 1-^nder F 

 (3-i3-> X A) we have the pressure in atmospheres which is ob- 

 tained by multiplying A by 3-132 (see Table I.) Under D we 

 have the sum of the expansions observed at each end '» te>™s 

 of the micrometer, divisions which had identical values, Under 

 F M'e have the values of D reduced to parts of an inch by multi- 

 plying- them by 0-000417. Under Q we have the greatest devia- 

 tions from the mean amongst the individual observations forming 

 this particular series. R represents this deviation as a percentage 

 of the total expansion (F). Under H we have the linear com- 

 pression (in inches) of a rod one million inches long under a 

 pressure of P atmospheres. K is the corresponding value for 

 one atmosphere, and N = 3K is the cubical compression of the 

 glass per million per atmosphere. , , , . . 



The total expansion on D was determined by observing the 

 expansion at each end and adding them^together. These partial 

 expansions were not always, nor indeed often, of exactly the 

 same extent ; the excess was sometimes on the one side and 

 sometimes on the other. The effect of the rise of pressure is to 

 extend the containing tube and to compress the contained rod 

 On the relief of pressure the tube shortens agam and the rod 

 recovers its length, and there is necessarily a sliding of the one 

 or the other, and 'it depends entirely on minute local circum- 

 stances whether the rod finds it easier to return to its origmal 

 relative position or to another. In some experiments made pre- 

 viously to those quoted in Table II. the rod had greater freedom 

 of motion longitudinally, and it happened several times that it 

 crept bodily to the one end, necessitating the opening of tlie 

 apparatus to replace it in a position suited to observation. Alter- 

 wards stops were placed in the tube, which, while setting limits 

 to the crawling motion, did not in any way interfere witli the 

 expansion and contraction. The results of these previous ex. 

 perimcnts are not included in the table, because they were merely 

 tentative with a view to learning the details of the kind of 

 experimentation ; and further, because in the microscope at the 

 east end the power used was very low, and the micrometer 

 insufficiently delicate. 



The micrometers used were : at the east end a photograptiic 

 copy of Hartnack's eyepiece micrometer, and at the west end 

 one of Morz's. They were both compared, and the values of 

 their divisions as used determined by comparison with a stage 

 micrometer of Smith and Beck, obligingly lent me by my friend 



■g M 



the deviation per cent, should be, as it is, the less the higher the 

 pressure; and there is no doubt that the higher the pressure is 

 the .greater is the accuracy of the observation. ^ The only way in 

 which the pressure affects the reading of the micrometer is hat 

 when it is sufficiently high it produces a microscopic distortion 

 of the tube which throws the point very slightly out of focus. 

 This is remedied by a slight touch .of the fine-adjustment screw 



"'^TlTe generaTresult of these experiments^ is that the linear com- 

 pressibility of the glass experimented on is 0-96 and its cubical 



"Gi:^''S-s^lJ^^i'S his observations at pressures up 

 totenatmosphere^s_: ,-804 and 2-S584 



So tlmt our results agree closely with those found by him for 

 crystal. _ ^^^^^^ ^^^.^^^_ ^^^^,^ _ ^^^^^^ j^^^ ^^^ p ^y^. 



