252 Mr. W. H. Dines. [June 19, 



; No marked change in the distribution of the stream lines occurred 

 on either the back or front of the plate at angles between 45 and 

 60, several intermediate positions having been tried. 



Until the angle of incidence exceeds 60 there is a region of still 

 air at the back just behind the forward edge of the plate; the 

 width of this seems to increase, but it remains at all events within 

 2 inches of the edge until the angle of incidence exceeds 60. It is 

 curious that at 75 the air should be at rest at the centre of the back 

 of the plate, and pass off laterally in both directions. In all cases 

 the motion on the front of the plate was quite steady. At the back, 

 iii the case of normal exposure, it was too unsteady to be deter- 

 mined ; at 15 it was decidedly unsteady ; but in the other positions 

 there was no difficulty in seeing the position of the silk threads. 



Effect of Temperature, Pressure, 8fc. 



It will be seen on reference to the tables that very divergent results 

 were sometimes obtained. Obviously experiments made under the 

 same conditions of barometrical pressure, temperature, and wind 

 Ought to give the same values. In cases where a small change of in- 

 clination causes a considerable change in the moment, a discrepancy 

 in the results is easily explained by supposing a small mistake to 

 have been made in measuring the angle, but in many cases this ex- 

 planation does not apply. Some natural wind ought to increase 

 the mean pressure, but in so far as I can judge it does not do so 

 to anything like the theoretical amount, that is to say, the pressure 

 is not increased so much as the mean square of the relative velocity. 

 Repeated experiments have shown that it is not possible to get a 

 different result by altering the rate of the whirling machine, all 

 Speeds under 70 miles an hour, which is the greatest of which the 

 machine is capable, giving results which show that the pressure 

 varies as the square of the velocity. It must, however, be remem- 

 bered that at low speeds the forces are too small to be capable of 

 exact measurement. 



Barometrical pressure has the result which might be expected, the 

 pressure on the plate varying directly as the height of the mer- 

 dury. A rise of temperature does not seem to make much differ- 

 ence, but, if anything, it increases the pressure. Experiments have 

 been made through a range of about 40 F., from 28 F. to 68 F. 

 The greater viscosity, I suppose, at the higher temperatures more 

 than compensates for the decrease of density, for certainly, other cir- 

 cumstances being the same, the pressure is not less at 60 F. than at 

 30 F., and the lowest values ever obtained were in a thick fog with a 

 temperature below the freezing point. 



No determinations of the dew point have ever been made in con- 



