ON THE MEASUREMENT OF THE LUNAR DISTURBANCE OF GRAVITY. 1 1 7 



gether. At the suggestion of Leverrier the iron basin -was replaced by 

 a shallow wooden tray with a corrugated bottom, and a good reflection 

 was then generally obtainable. Immediately over the mercury pool 

 there stood a lens of 10 cm. diameter and 10 meters focal length, 

 and over the brass disk there stood a microscope with moveable mi- 

 crometer wires in the eyepiece, and a jaosition circle. The platinum 

 wires were illuminated, and on looking through the microscope the 

 observer saw the wires both directly and by reflection. The observations 

 were taken by measuring the azimuth and displacement of the image of 

 the central square relatively to the real square enclosed by the wires. 



One division of the micrometer screw indicated a displacement of 

 vertical of 0"'03, so that the observations were susceptible of consider- 

 able refinement. 



The whole of the masonry was finished in 1863, and M. d'Abbadie 

 then allowed the structure five years to settle before he began taking 

 observations. The arrangements for observing above described were 

 made in 1868 and 1869. 



In the course of a year he secured 2,000 observations, and the results 

 appear to be very strange and capricious. 



Throughout March, 1869, the perturbations of the mercury were so 

 incessant, that observations (taken at that time with the iron basin) were 

 nearly impossible ; on the 29th he waited nearly an hour in vain in trying 

 to catch the image of the wires. Two days later the mercury was 

 perfectly tranquil. On April 6 it was much agitated, although the air 

 and sea were calm. A tranquil surface was a rare exception. 



In 1870 the corrugated trough was substituted for the iron basin ; 

 and M. d'Abbadie says : — 



' Cependant, ni le fond inegal du bain rainc ni sa forme ne m'ont 

 empeche d'observer, ce que j'appelle des ombres fuyantes. Ce sent des 

 bandes sombres et paralleles qui traversent le champ du microscope avec 

 plus ou moins de vitesse, et qu'on explique en attribuant au mercure des 

 ondes tres tenues, causees par une oscillation du sol dans tin seul sens. 

 Le plus souvent ces ombres semblent courir du S.E. au N.O., appi'oxi- 

 mativement selon I'axe de la chaine des Pyrenees ; mais je les ai obser- 

 vees, le 15 Mars 1872, allant vers Ic S.O. A cette epoque le mercm^e 

 etait, depuis le 29 fevrier, dans une agitation continuelle, comme mon 

 aide I'avait constate en 1869, aussi dans le mois de Mars.'' 



He observed also, from time to time, certain oscillations of the 

 mercury too rapid to be counted, which he calls ' tremoussements.' 

 There were also sudden jumpings of the image from one point to another, 

 or ' fretillements,' indicating a sudden change of vertical through 0"'49 

 to 0"-65. 



He observed many microscopic earthquakes, and in some cases the 

 image was carried quite out of the field of view. 



He also detected the difference of vertical according to the- state 

 of the tide in the neighbouring sea ; but the change of level due to this 

 cause was often masked by others occurring contemporaneously. 



From observations during the years 1867 to 1872 (with the exception 

 of 1870) he finds that in every year but one the plumb-line deviated 

 northwards during the latter months of the year, but in 1872 it deviated 

 to the south. 



' M. d'Abbadie writes to me that this phenomenon was ultimately found to 

 result from air-currents (Nov. 6, 1881). 



