252 Dr. J. S. Gr. Thomas on Thermal Effect of a slow 



velocities of the impressed stream of air or other gas. A 

 number of results employing various members of the series 

 of wires for this purpose are given in figs. 6 and 7. 



The deflexion-velocity calibration curves given in fig. 6 

 were obtained by inserting various members, or groups of 

 members, of the wires adjacent to one another in opposite 

 arms of a Wheatstone bridge, as explained in previous 

 papers. The remaining wires were not heated, so that the 

 temperature conditions of the various wires were not those 

 indicated in fig. 1. The constant ratio arm was throughout 

 adjusted to a constant value of 1000 ohms, and the sensitivity 

 of the galvanometer w^as maintained constant throughout, 

 the galvanometer being shunted with 10 ohms. The current 

 employed was 1*1 amp. The curve marked 1 was obtained 

 by inserting the first pair of wires, one in each appropriate 

 arm of the bridge. The hot-wire anemometer so consti- 

 tuted is a simple hot-wire directional instrument similar 

 to that described in a previous paper. It has been shown 

 that the sensitivity of such an instrument is, in the region 

 of low velocities, considerably greater than that of the 

 corresponding Morris type of instrument. The curve marked 

 2 is the calibration curve obtained when the wares 1 and 2 

 are inserted in one arm of the bridge, and wires 3 and 4 

 in the opposite arm. A greater sensitivity of the bridge 

 arrangement is to be anticipated for the latter arrangement, 

 as, employing the same total current in the bridge with 

 zero flow, the out-of-balance current through the galvano- 

 meter is approximately doubled on doubling the resistances 

 in the respective arms of the bridge. Except in the neigh- 

 bourhood of the origin, it will be seen that the sensitivity 

 of the arrangement employing two wires, as described, 

 in each arm of the bridge is considerably more than 

 twice that when one wire alone is inserted in the re- 

 spective arms. A similar result was found in the cases 

 where 3, 4, or 5 wires, made up of an appropriate number of 

 consecutive wires, were inserted in each arm of the bridge. 

 Thus from fig. 6 it will be seen that, corresponding to the 

 use of 1, 2, 3, 4, and 5 wires in the appropriate arms of the 

 bridge, the deflexions corresponding to a mean impressed 

 velocity of the air stream equal to 8 cms. per sec. were re- 

 spectively 57 , 185, 355, 485, and 527 divisions respectively. 

 The results are readily explainable from a consideration of 

 the increase or decrease of the temperature of the respective 

 wires corresponding to an impressed velocity of 8 cms. per 

 sec. Thus, in the case where all the wires are heated, under 

 the influence of this impressed velocity, as seen in fig. 2, the 



