Febeuakt 22, 1907] 



SCIENCE 



303 



SOCIETIES AND ACADEMIES 



THE PHILOSOPHICAL SOCIETY OF WASHINGTON 



The 626tli regular meeting was held Jan- 

 uary 5, 1907, with President Hayford in the 

 chair. 



The first paper of the evening was presented 

 by Mr. L. W. Austin, describing some recent 

 developments of wireless telegraphy and giv- 

 ing a comparison of the efficiency of continu- 

 ous and broken wave trains in wireless signal- 

 ing. 



The experiments were carried on at the 

 wireless station at Brant Eock, Mass., during 

 the past autumn, the distance between the two 

 antennae being about three miles. The con- 

 tinuous wave trains were produced by a small 

 Fessenden high frequency dynamo which dur- 

 ing the experiments was run at 50,000 cycles 

 per second. The broken wave trains were of 

 the same frequency and produced by a spark 

 in the usual manner. The receiver used was 

 of the electrolytic type. Below are given the 

 data of one of the experiments : 



Machine Spark 



Energy 50 watts 225 watts 



Current in sending 



antenna 1.6 amp. 0.9 amp. 



Strength of signal 5 20 ^ 



Current in receiving I '"^ arbitrary 



antennae 2.2 4.5 ) ™^*® 



the loudness of the signal being proportional 

 to the square of the received current. If we 

 reduce the received current to terms of the 

 same radiation current, the spark would give 

 3.6 times as strong received current as the 

 machine. Eeduced to terms of the same en- 

 ergy the two are nearly equal in efficiency, 

 the advantage being slightly in favor of the 

 machine. 



Mr. O. B. French spoke of ' The Eecent Use 

 of Invar Tapes for the Measurement of Pri- 

 mary Bases.' 



Since the discovery of the alloy of nickel 

 and steel (called invar, from invariable) which 

 possesses a very small coefficient of expansion, 

 its use for precise measuring apparatus has 

 been tested very carefully. Most of these in- 

 vestigations have been made under the direc- 

 tion of C. E. Guillaume, of the International 



Bureau of Weights and Measures at Paris. 

 His experiments having proved the metal to 

 be fairly stable, the Coast and Geodetic Sur- 

 vey decided to try it for the measurement of 

 primary base lines. 



In December, 1905, the survey purchased 

 from J. Agar Baugh, London, England, sev- 

 eral ribbons of the invar tapes, 6.3 mm. in 

 width, 0.5 mm. in thickness and 53 meters in 

 length, which were prepared for measuring 

 tapes in the instrument division of the survey. 



During 1906 the survey measured six base 

 lines, using on each base three invar tapes, in 

 daylight (standardized at the National Bureau 

 of Standards) and also three steel tapes, at 

 night (standardized in the field). 



Several pieces of the invar tapes, tested at 

 the Bureau of Standards, showed a tensile 

 strength of 100,000 pounds per square inch 

 (about one half that of steel tapes), with the 

 elastic yield point about 70 per cent, of the 

 tensile strength. 



The tapes were tested for considerable 

 ranges of temperature, reeled and unreeled a 

 large number of times, and also tested for 

 continued application and removal of light 

 loads, without showing any change in length. 

 The coefficient of expansion of the invar tapes 

 was found to be .0000004 per degree centi- 

 grade or 1/28 that of steel. 



The steel and invar measures of the six 

 bases were computed independently. The dif- 

 ferences between them are small, the largest 

 being 1:300,000 (3 mm. per km.) and the av- 

 erage about 1 :500,000. 



The probable errors of the lengths of the 

 bases from the steel measures are more than 

 double those from the invar measures. The 

 final probable errors of the bases, giving the 

 invar double weight, are between 1:2,500,000, 

 and 1:5,000,000 (0.4 mm. and 0.2 mm. per 

 km.). 



In 1900 the Coast and Geodetic Survey 

 demonstrated that steel tapes gave practically 

 the same accuracy as bar apparatus with one 

 third of the cost. It is now shown that the 

 invar tapes give results considerably more 

 accurate and economical than the steel tapes. 



Mr. W. P. White then made some informal 

 remarks upon ' Suspended Galvanometer Sup- 



