On the Thermal Properties of Normal Pentane. 353 



began to run about over the surface, spinning the fine powder 

 up in funnel-shaped vortices, which lasted sometimes ten or 

 fifteen seconds. The silica powder must be made by igniting 

 the gelatinous precipitate formed when silicon tetrafluoride is 

 conducted into water. The commercial article is not suffi- 

 ciently light and mobile. 



Whirls formed in this way cannot be seen by a large 

 audience, however, and I accordingly sought some way of 

 making them on a larger scale. The plate was well heated 

 after removing the silica and then dusted with sal-ammoniac: 

 dense clouds of white vapour immediately arose from the hot 

 surface, and presently in the centre there mounted to a height 

 of about 2 metres a most perfect miniature tornado of dense 

 smoke. By placing the plate in the beam of a lantern in a 

 dark room, the whirls can be shown to a class in a large 

 lecture-room. I find that it is best to put on the sal-ammoniac 

 first and then heat the plate: the vortices then come off the 

 plate almost continuously, and often persist for some time. 



An instantaneous photograph of one of these tornadoes was 

 taken in bright sunshine, and is reproduced in fig. 7. 



This method of showing atmospheric vortices seems far 

 preferable to the old way of forming them, by means of a 

 rapidly whirling drum with cross partitions, as the whirls are 

 produced by the same causes and under the same conditions 

 that they are in nature. 



Physical Laboratory 



of the University of "Wisconsin, 



Madison, Nov. 20. 



XXXI. On the Thermal Properties of Normal Pentane. By 

 J. Kose-Innes, M.A., B.Sc, and Sydney Young, D.Sc, 

 F.R.S* 



IX the year 1894 an experimental investigation of the 

 relations between the temperatures, pressures, and 

 volumes of Isopentane, through a very wide range of volume, 

 was carried out by one of us, and the results were published 

 in the Proc. Phys. Soc. xiii. pp. 602—657. It was there shown 

 that the relation p = bT — a at constant volume (where a and 

 b are constants depending on the nature of the substance and 

 on the volume) holds good with at any rate but small error 

 from the largest volume (4000 cub. cms. per gram) to the 

 smallest (1*58 cub. cms. per gram). 



In the neighbourhood of the critical volume (4*266 cub. 

 cms.), and at large and very small volumes, the observed 

 deviations were well within the limits of experimental error 

 * Communicated by the Physical Society: read December 9, 1898. 



