June 9, 1911] 



SCIENCE 



911 



In use the tank is filled with alcohol or non- 

 freezing liquid (about 3i galloms), and the flask 

 charged (two thirds or more) with liquid ammonia 

 (5 to 6 pounds). To lower the temperature, 

 gaseous ammonia is allowed simply to escape 

 slowly from the flask. The fumes may be led 

 into a vessel of water with the formation of 

 aqueous ammonia, or allowed to escape entirely. 

 A strong circulation of the liquid of the bath is 

 maintained by means of an efficient form of 

 motor-driven stirrer. 



The temperature can be lowered to —25° to 



— 30° C, by the spontaneous evaporation and 

 escape of the ammonia. For still lower tempera- 

 tures it is necessary to use a suction pump to 

 increase the evaporation by reduction of pressure. 

 A simple hand pump was described for this pur- 

 pose with the novel feature that the outlet valve 

 opened directly into a water jacket space sup- 

 plied with running water. Moreover, the valve 

 itself had a small hole directly through it, so that 

 water constantly leaked into the pump chamber 

 and there absorbed large volumes of gaseous 

 ammonia. When in operation, the object of 

 working the piston of the pump up and down is 

 quite as much to expel the water leaking into 

 the pump chamber as to take off the gas, large 

 quantities of which are absorbed by the fresh 

 water entering the pump after each stroke. 



An apparatus of this character has been used 

 for many years at the Weather Bureau in the 

 comparison of thermometers and other low tem- 

 perature work, and was first described in Annual 

 Eeport of the Chief Signal Officer, 1891-92, 

 p. 355. The bath can readily be lowered to 



— 40° C, and thermometer comparisons carried 

 on over a period of four or five hours with an 

 expenditure of not over five or six pounds of 

 ammonia, which costs about $1.50. Experience 

 demonstrates that in a small apparatus of the 

 kind described, it is much more convenient and 

 economical to let the ammonia escape after evap- 

 oration than to try to recondense it back to the 

 liquid state. It proves to be quite practicable to 

 carry the temperature down to — 70° C, at which 

 a number of thermometer comparisons have al- 

 ready been made. 



Becent Gravity Work hy the Coast and Geodetic 

 Survey: Wm. Bowie, of the Coast and Geodetic 

 Survey. 



This paper gave an account of the gravity 

 work done by the Coast and Geodetic Survey 

 during the past few years. In 1891 the survey 

 began the use of two sets of half -second pendu- 



lums which proved efficient and accurate in this 

 relative determination of gravitation, using the 

 gravity pier at the Coast and Geodetic Survey 

 Office as the base station. As first constructed 

 the knife edges were fastened to the head of the 

 pendulum with the supporting planes on the pen- 

 dulum case. Several years later this method was 

 reversed, the planes being placed in the pendulum 

 head and the knife edges on the case. This gave 

 a more invariable length of the pendulum, as the 

 effect of any wear on the planes would be neg- 

 ligible. 



The different apparatus used in making the 

 observations were illustrated by lantern slides. 

 Previous to 1909 the relative value of the in- 

 tensity of gravity had been determined at forty- 

 seven stations in the United States. In that year 

 a campaign of gravity work was begun and is still 

 in progress. It will probably close at the end of 

 the present year. During the past two and one 

 half years fifty-six stations have been established. 

 By the end of this year the intensity of gravity 

 will have been determined at about twenty addi- 

 tional stations. This will make a total in the 

 United States of about one hundred and twenty 

 stations. 



The apparatus and methods are sensibly the 

 same as those used in previous work except that 

 an interferometer was used for determining the 

 flexure of the pendulum case due to the horizontal 

 force applied by the swinging pendulum. This 

 took the place of the static method. The inter- 

 ferometer had never been used previously in any 

 country for determining flexure on the field. 



In 1909 J. F. Hayford, the inspector of geodetic 

 work in the Coast and Geodetic Survey, made an 

 investigation of the effect of topography and iso- 

 static compensation on the intensity of gravity, 

 using fifty-six stations in the United States. He 

 read a paper showing his results at the meeting 

 of the International Geodetic Association at Cam- 

 bridge and London in that year. By the applica- 

 tion of the theory of isostasy the large anomalies 

 between the theoretical and actual value of gravi- 

 tation by the Bouguer and free air reductions 

 were greatly reduced. The mean anomalies with- 

 out regard to sign for the fifty-six stations were 

 .064, .027 and .015 dyne for the Bouguer, free 

 air and Hayford methods, respectively. 



An analysis of the anomalies by the new method 

 showed that there is only a slight, if any, connec- 

 tion between the size and sign of the anomaly and 

 the character of the topography surrounding the 

 station. 



