WITHDRAWN FROM THE ACTION OF GRAVITY. 431 



determine the capillary height, the process of Gay-Lussac was employed, that 

 is to say, the measurement by the cathetometer, all known precautions being 

 taken to secure an exact result. The experiment was made at the temperature 

 of 19°. I had procured a capillary tube having an interior diameter of but a 

 fi-action of a millimetre; for what reason will presently be seen. First, a slight 

 mark was made with a file upon this tube at about 3| centimetres from one of 

 its extremities, a distance which had been found, by previous trial with another 

 fragment of the same tube, to be a little greater than the height of the capillary 

 column raised ; next, the interior of the tube was thoroughly moistened by 



f dunging it several times to the bottom of the vessel containing the glyceric 

 iquid and shaking it each time on withdrawing it. Lastly, after having wiped 

 it outside, it was put in place by immersing it in the liquid until the extremity 

 of the column raised appeared to stop very near the mark, and the point of steel 

 was lowered so as to be on a level with the exterior liquid. The horizontal 

 thread of the eye-glass of the cathetometer was then brought into contact with 

 the image of the lowest point of the concave meniscus, and at intervals of five 

 minutes this contact was verified and re-established until the point in question 

 appeared stationary ; nor yet Avas measurement made until, by the lapse of half 

 an hour, the perfect immobility of the summit of the column was ascertained. 

 The movements had been very small, so that the column still terminated near 

 the mark. The reading at the cathetometer gave, for the distance from the 

 lowest point of the concave meniscus to the exterior level, 27'^^.35. 



This measurement having been taken, the tube was removed, cut at the mark, 

 and its interior diameter at that point measured by means of a microscope fur- 

 nished with a micrometer giving directly hundredths of a millimetre. It waa 

 found that the interior section of the tube was slightly elliptical, the greater di- 

 ameter being 0™"'.374, and the smaller 0'"'".357; the mean was adopted, 

 namely, 0™"^.3655, to represent the interior diameter of the tube assumed to be 

 cylindrical. To have the true height of the capillary column, it is necessary, 

 we know, to add to the height of the lowest point of the meniscus the sixth part 

 of the diameter of the tube, or, in the present case, 0"'"^.06 ; the true height of 

 our column is consequently 27™".41. Now, to obtain the height h to which 

 the same liquid would rise in a tube having an interior diameter of exactly a 

 millimetre, it is sufficient, in virtue of the known law, to multiply the above 

 height by the diameter of the tube, and thus we find definitively A=10"^'".O18. 

 I should here say for what reason I have chosen for the experiment a tube 

 whose interior diameter is considerably less than a millimetre. The reasoning 

 by which 1 arrived (§ 23) at the formula supposes that the surface which ter- 

 minates the capillary column is hemispherical ; now that is not strictly true, 

 but in a tube so narrow as that which 1 have employed, the difference is wholly 

 imperceptible, so that in afterwards calculating, by the law of the inverse ratio 

 of the elevation to the diameter, the height for a tube one millimetre in diame- 

 ter, we would have this height such as it would be if the upper surface were 

 exactly hemispherical. 



The values of p and h being thus determined, we deduce therefrom 2lip=22.17, 

 a number which differs but little from 22. ."^G, obtained in the preceding para- 

 graph as the value of the product pd. The formula pd^=2/ip may therefore be 

 regarded as verified by experiment, and the verification will appear still more 

 complete if we consider that the two results are respectively deduced from ele- 

 ments altogether different. I hope hereafter to obtaui new verifications with 

 other liquids. 



