by clock-work, C, makes the plate move forwards and backwards. The forward movement brings about a momentary contact of the recording tip with the smoked plate inscribing a dot. These single dots are made at intervals of 15 minutes; at the expiration of the hour, however, contact is made three times in rapid succession, printing a thick dot. It is thus easy to determine the movement of the tree at all times of the day and night. A second lever, R, placed above, gives on the same plate, thermographic record of the diurnal variation of temperature. For this I use a differential thermometer, T, made of a compound strip of brass and steel. Curvature is induced by the differential expansion of the two pieces of metal. The up or down movement of the free end of the compound strip is further magnified by the recording lever. This arrangement was extremely sensitive and gave accurate record of variation of temperature. By the forward movement of the oscillating plate two dots are made at the same time,—one for the temperature and the other for the corresponding movement of the tree. As the two recorders do not move vertically up or down, but describe a circle, the dots vertically one above the other may not correspond as regards time. Any possibility of error in calculation is obviated by the fact that the thick dots in both the records are made every hour, and the subsequent thin dots at intervals of 15 minutes.
A difficulty arose at the beginning in obtaining sanction of the proprietor to attach the recorder to the tree. He was apprehensive that its miraculous power might disappear by profane contact with foreign-looking instruments. His misgivings were removed on the assurance that the instrument was made in my laboratory in India, and that it would be attached to the tree by one of my assistants, who was the son of a priest.
From results of observation it is found that the tree moves through its entire length; the fall of the highest
