466 Mr. W. N. Shaw. On a Pneumatic [Apr. 24, 



northward, although they had not been intentionally magnetised, 

 nor, indeed, exposed to any risk of their being so from the time of 

 their being purchased. They were, no doubt, very slightly magnetised, 

 but the time of swing was very long, and the position of equilibrium 

 not sufficiently definite. I therefore magnetised them more strongly \ 

 the little vane then took up, in consequence, a definite position of 

 equilibrium with the planes of the vanes approximately north and 

 south. The apparatus being so placed that the tube, G, is east 

 and west, the vanes always set across the tube when there is no 

 current. The needle points enable the position of equilibrium to be 

 clearly identified by the aid of a fiducial mark on the glass tube. 

 The sensitiveness can be altered as desired by an external control 

 magnet, just as that of a galvanometer needle can be. The little 

 compass needle or wind vane, M, is very sensitive to the motion of 

 air in the tube, and although it may be possible to find other detectors 

 that are equal, or even superior to it, yet the ease of seeing it, the 

 rapidity of its action, and its definite zero are decidedly in its favour. 



The head is produced by a gas burner in a metal chimney, C, fitted 

 to the lid of the box A. 



Various precautions are required in fitting the boxes together to 

 secure that the air should only flow through apertures intended for 

 its passage, but they need not here be detailed. They consist mainly 

 in the plentiful application of glue and brown paper. 



The apparatus was designed when I was making a number of 

 observations of flow of air to illustrate the theory of ventilation, and 

 I did not anticipate that any high degree of accuracy could be aimed 

 at. I was, therefore, agreeably surprised to find that the identification 

 of the condition of no flow is capable of much greater accuracy than 

 the arrangements for measuring the areas of the orifices would allow 

 me to interpret. 



Of the four apertures of the bridge, two, viz., a 2 and a 4 , are in- 

 accessible without pulling the arrangement to pieces ; they represent 

 areas of -J sq. in. and 1 sq. in., respectively, as accurately as a knife 

 could cut them in cardboard. 



The other two areas, viz., and a 3 , are made by sliding shutters, 

 as already mentioned. Their edges were cut with a knife, and they 

 probably are only rough approximations to areas in a truly thin plate,, 

 so that little importance can be attached to the final results of the 

 measurements which will be given below ; they serve only to show 

 that the width of the adjustable slit, when there is no flow through 

 the galvanometer tube, is a perfectly definite magnitude. 



The following observations have been taken with the apparatus : — 



