Februaey 6, 1903.] 



SCIENCE. 



217 



calculated therefrom equals the known 

 velocity of the air; (4) the ditt'erential 

 pressure varies exactly as the square of 

 the velocity, as required by theorJ^ The 

 air velocities employed ranged from five 

 to thirty miles an hour, and the pressure 

 gauge was graduated to millionths of an 

 atmosphere. The experiments were con- 

 ducted in a tunnel through which air was 

 drawn with uniform velocity and direction, 

 its velocity being measured simultaneously 

 by the pressure-tube anemometer and by 

 a balloon anemometer. In the latter 

 device a toy balloon drifting through the 

 tunnel cuts two pencils of light thrown 

 squarely across its path at an interval of 

 ten feet, the time between the cutting of 

 the sheets of light being determined pho- 

 tographically. The average wind-speed 

 determined by means of the pressure-tube 

 anemometer agrees with the average de- 

 termined by the standard, or balloon ane- 

 mometer, accurately to less than one per 

 cent. Examples are also given of the use 

 of the pressure-gauge for measuring static 

 pressures from one millionth of an atmos- 

 phere upwards. 



Hydrographic Work of the JJ. S. Geolog- 

 ical Survey: H. A. Pressey, "Washing- 

 ton, D. C. 



The work of the Survey in measuring 

 the flow of all the important streams in 

 the country is described, and the great 

 value of the results to industrial projects 

 pointed out. 



Friction in Ball-hearings : M. J. Golden, 

 Purdue University, Lafayette, Ind. 

 The paper describes an apparatus used 

 to determine the friction of ball-bearings 

 of different sizes at different speeds. It 

 was shown that at high speeds ball-bear- 

 ings fail entirely. Ball-bearings for ordi- 

 nary pressures and speeds give a loss by 

 friction less than that of an ordinary bear- 

 ing poorly lubricated, but not much less 



than a finely polished and thoroughly lu- 

 bricated bearing. 



Errors in Analyses of Furnace Gases 

 Shoivn by Computation: William Kent, 

 Passaic, N. J. 



It is shown by arithmetical computation 

 based on the analyses of a certain coal that 

 the analysis of the gas from the chimney, 

 as reported by the chemist, must be in 

 error. "With such an analysis it is impos- 

 sible to compute a heat balance in a boiler 

 test with any approach to accuracy. 



Heat Exchanges Witliin the Steam-engine: 

 R. H. Thurston, Cornell University, 

 Ithaca, N. Y. (Not read.) 

 The method of heat exchange in the 

 steam-engine cylinder, which results in 

 serious wastes of heat and proportional 

 reduction of the efficiency of the machine, 

 has been considered an obscure phenom- 

 enon. The experiments made by Professor 

 Dwelshauvers-Dery and by M. Duchesne 

 indicate that the cylinder wall takes the 

 temperature of the steam as long as it is 

 covered with moisture; but when the wall 

 is dry it may hold a temperature consid- 

 erably in excess of that of the steam in 

 contact with it. During expansion and 

 compression of steam in the cylinder there 

 is a constant interchange of heat, which 

 accounts for the varying efSciency of the 

 steam as a motor. Experiments con- 

 ducted at Sibley College of Cornell Uni- 

 versity sustain these deductions. The ex- 

 periment is described and results .shown 

 graphically. Elwood Mead, 



Secretary. 



SECTION E, GEOLOGY AND GEOGRAPHY. 

 Some forty-five papers were offered to 

 Section E for reading at the Washington 

 meeting. On account, however, of the con- 

 flict with the meeting of the Geological 

 Society of America, all the papers of the 

 Section E program were accepted by the 



