450 report — 1878. 



and 15 inches long, in which a piston slides so as to regulate the length of the 

 tube. This tube is fixed in a block of wood, to which is attached a tuning- 

 fork, whose points are just above the open end of the tube. The tuning-fork is 

 sounded in any convenient way, and the piston is moved out and in, until the 

 proper length is found, which is indicated by the resonator intensifying the 

 sound of the tuning-fork. With practice the length can thus be determined 

 with an accuracy of at least 1 in 250. But the length of resonator depends 

 upon the density of the gas, a light gas requiring a longer resonator; and by 

 reading off on a scale the position of the piston we judge of its density. In 

 this manner 1 or 2 per cent, of fire-damp mixed with common air can be detected. 

 Barometric pressure produces no difference. The temperature correction is made 

 by reading off a thermometer of the proper dimensions in place of reading off a 

 fixed mark on the piston. The only error is by the presence of dense carbonic 

 acid gas. But carbonic acid gas tends to destroy the explosive character of fire- 

 damp, and it appears that if the presence of carbonic acid prevented the instru- 

 ment from indicating fire-damp it would certainly be sufficient to prevent the 

 explosive character of the fire-damp. 



7. On Certain Phenomena Accompanying Rainbows* 

 By Professor Silvanus P. Thompson, D.Sc, B.A. 



The author narrated several instances of rainbows, seen chiefly in Switzerland, 

 where radial streaks of light devoid of colour were observed within the primary 

 and without the secondary bow. The explanation suggested was as follows : The 

 wedge-shaped radial streaks are beams of sunlight, which become visible by diffuse 

 reflection from particles of matter in their path, just as the apparently divergent 

 beams of sunrise or sunset become visible. These " beams " being practically parallel 

 to one another, appear to converge in the point exactly opposite the sun by perspec- 

 tive ; or, in fact, just as the parallel beams of sunset appear divergent. Since the 

 rainbow has for its centre the point opposite the sun, such beams must have posi- 

 tions radial with respect to the bow. They resemble, therefore, the Rayons du 

 Crepuscule occasionally seen in the east at sunset. They had never been observed 

 crossing the dark space between the primary and secondary bows. A similar 

 phenomenon of rays might be sometimes seen in sunlight when the shadow of the 

 observer fell upon a slightly turbid lake or river. 



8. New Magnetic Figures. By Silvanus P. Thompson, B.A., D.Sc, 

 Professor of Experimental Physics in University College, Bristol. 



The author desires to draw the attention of the Section to the series of magnetic 

 figures exhibited, which has recently been completed. The figures are permanently 

 secured on glass by a process described in recent communications to the Physical 

 Societies of London and Paris ; and they have also been photographed as trans- 

 parencies for the lantern. 



It is believed that the figures assumed by iron filings in magnetic fields of many 

 different kinds, in the series now produced, will be found of great use in the study 

 and teaching of known laws of magnetic and electrodynamic action, and also in 

 the experimental determination of the action of magneto-dynamic and electro- 

 dynamic systems and machines. Faraday discovered that the lines of force traced 

 out thus by iron filings possessed a significance hitherto disregarded, and revealed 

 indeed the very seat of the attractions and repulsions taking place between 

 magnetic bodies. The method had indeed been imperfectly employed by Mus- 

 schenbrock at an earlier date, but it only became really fruitful in the hands of 

 Faraday. The author has applied Faraday's reasoning to the figures produced by 



* < 



Philosophical Magazine,' October 1878. 



