OF ARTS AND SCIENCES. 327 



inclination to the horizon. Beneath there is a layer of sand. The 

 tremors communicated to the clock pier through this combination were 

 found to be surprisingly great. The tremors produced by ice-carts at 

 the distance of 1,000 feet rendered it impossible to make exact com- 

 parisons. The heavy steps of an assistant could be counted to the 

 distance of 100 feet, by noticing the effect of the concussion upon the 

 pier, which was in turn communicated to the surface of the bar ex- 

 amined under the microscope. 



Heat is communicated to this comparing-room from the furnace in 

 the cellar by a pipe which enters the room near the ceiling and nearly 

 over the comparator. Under certain circumstances the temperature 

 in this room may be kept under very good control, but great care has 

 been found necessary in this regard. The comparator is directly in 

 the line of the windows. By opening both windows during the night, 

 the temperature within the room is found to be reduced nearly to the 

 temperature of the outside air. By making the observations with this 

 comparator on cloudy days and early in the morning, much better re- 

 sults have been obtained than were expected. 



By the liberality of the President of Harvard College, comparing- 

 room No. 2 was fitted for the reception of the Universal Comparator. 

 It is situated in the basement of Harvard Hall. The dimensions of 

 the room are 12 feet in length, 9 feet in width, and 8 feet in height. 

 The brick walls which surround the room are twelve inches in thick- 

 ness. The brick piers upon which the comparator rests extend to a 

 depth of eight feet. The walls, the ceiling, and the floor are all double- 

 planked, with two intervening air spaces. Between the inner and the 

 outer partition there is a layer of rosin-soaked paper. The room has 

 double windows and double doors. The steadiness of the tempera- 

 ture within the room may be inferred from the record on pp. 366-371. 

 Durino; the summer of 1882 the extreme change of the Fahrenheit 

 thermometer for eicjht weeks was 1°.3. During the same time the 

 daily variation in the temperature of the outside air often amounted 

 to 25°. 



Between April 14 and June 1, 1883, the extreme range was 0°.75 C. 

 On June 3 the temperature of the room was raised by artificial means. 

 By June 5 the temperature had reached the stationary point, viz. 

 17°. 07 C, and between this date and June 28 the rise was only 0°.96. 

 On July 6, the reading of 1^61 was only 18°. 48, notwithstanding the 

 average temperature of 31° for the two preceding days. On July 7, 

 the room was exposed to the open air. 



An attempt has been made to produce desired variations of temper- 



