May 19, 1892] 



NA TURE 



55 



interest, summer resorts, fishing places, &c., in Eastern 

 Ontario, the Muskoka district, the St. Lawrence region, 

 the Lake St. John country, the maritime provinces, Prince 

 Edward Island, and Newfoundland. In an appendix 

 are given fish and game laws, and official lists of trout 

 and salmon rivers and their lessees. The author gener- 

 ally compresses his information into as small a space 

 as possible, but in dealing with the more interesting 

 Canadian scenes has sought to make his descriptions 

 lively and attractive. The volume is prettily printed, 

 and is well supplied with maps and illustrations. 



LETTERS TO THE EDITOR 



[The Editor does not hold himself responsible for opinions ex- 

 pressed by his correspondents. Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part of ^\T\JVit.. 

 No notice is taken of anonymous communications. "[ 



A Question in Physics. 



Can there be a crowding of the particles of a gas to a much 

 smaller compass without its being markedly heated? Can a 

 gas expand without being cooled ? It is probable that nearly 

 every physicist will give negative answers to these questions, 

 and yet the fact that such conditions may occur sometimes 

 seems well established. The present writer, in 1889, attempted 

 to determine the actual heating of air when compressed by a 

 pump connected with the cylinder by a long tube, and found 

 that the temperature was raised about 4° F. for a compression of 

 10 inches above atmospheric pressure. In like manner, on ex- 

 panding this compressed air into the free atmosphere, it was found 

 that the cooliag was about 4^ These results were published in 

 Science, vol. xv. p. 387, and were strongly combated by Prof 

 Ferrel and Prof Marvin. Prof. Ferrel advanced, as applic- 

 able in this case, the well-known thermodynamic formula for 

 the computation of the heat developed in a gas when compressed, 

 as follows : — 



hii) 



in which t and t' are the absolute temperatures corresponding to 

 the pressures p and />'. Sir Wm. Thomson has given this 

 formula in slightly different form, and with a larger exponent 

 (see " Encyclopaedia Britannica," vol. vii. p. 814). Prof Ferrel 

 found that, under the experimental conditions above, the heating 

 should have been 43°, and the cooling 45°? (38°) (see Ame- 

 rican Meteorological Journal, vol. xii. pp. 339 and 340). 



It seems very evident, however, that this formula can be used 

 only when all the heat due to the work of compression is con- 

 centrated in the compressed air, and conversely when the air 

 expands against an external resistance. An experiment by 

 Joule will serve to elucidate this point. He determined the 

 mechanical equivalent of heat by immersing the cylinder into 

 which the air was to be compressed and the compressing pump in 

 the same water-bath, and then determining the amount of com- 

 pression and the total heat developed. This enables us to 

 advance the proposition : If air when compressed is to be raised to 

 the temperature indicated by theory, it is very essential that all 

 the heat developed in the work of compression should enter the air. 

 This seems self-evident ; nevertheless, nearly all the errors that 

 have entered the various discussions of this question have arisen 

 from a neglect of this very obvious consideration. 



In Joule's experiment let us suppose that the compressing 

 pump had been in one bath, and the cylinder into which the air 

 was compressed in another. Under these conditions, if no heat 

 were lost, the first bath would have received very much the 

 greater amount of heat. Now, if the compressed air in passing 

 from the pump to the cylinder became cooled to the outside 

 temperature, it is evident that all the heat due to the work of 

 compression would have been dispo-;ed of outside the cylinder, 

 and would not have been available for raising the temperature 

 of the compressed air. 



Instead of connecting the pump directly with the cylinder, 

 let us take two cylinders of the same size, and connected by 

 a tube. Compress the air in the first cylinder (A), to three 

 atmospheres, the air in the other (B) being at atmospheric 

 pressure. If we cool the air in A to the outside temperature, 



NO 



II 77, VOL. 46] 



and then open the connection with B, the compressed air will 

 rush from A to B, and an equilibrum will be established 

 very quickly, the pressure in each cylinder being at two 

 atmospheres. The air in A will be slightly chilled because 

 of the work of imparting a certain velocity to the particles 

 entering B, and the air in B will be slightly warmed from 

 the impact of the particles rushing from A, but there will be 

 no healing due to the work of an external force making the 

 compression. 



Instead of allowing the air in A to rush into B, suppose 

 we open communication with the outside air. The resistance 

 to the rush of air will be much less than before, and the chilling 

 in A, due to the work of imparting a certain velocity to the 

 air, would be slightly greater than in the previous case, but it is 

 obvious that this will be vastly less than that given by the 

 formula. We may say, then, that the conditions suggested by 

 the questions above may be very easily brought about. 



The compressed air in a cylinder has a potential energy or 

 capacity to do work, and this energy may be transmitted to 

 another cylinder having air at atmospheric pressure without loss, 

 and plainly without imparting or losing any heat. We might 

 compare it to the head of water in a pond. This water has 

 a certain capacity to do work depending upon its head. We 

 may enlarge the pond somewhat, and the capachy for doing 

 work will remain almost unchanged. The extremely important 

 bearing of these views upon problems in meteorology is very 

 apparent. The convection theory of storms demands a cooling 

 from the work of expansion in an ascending column of moist 

 warm air ; it would appear, however, that the cooling must be 

 vastly less than has generally been considered probable. 



H. A. Hazen. 



Aurora. 



Perhaps it may interest some of your readers to see a short 

 abstract of the observations of aurora made here during the last 

 months, this winter having been by far the richest in well- 

 developed northern lights since the winter 1870-71. Beginning 

 with the magnificent display of February 13, which lasted 

 almost the whole night, sometimes with vivid red and green 

 tints (it was first noted at 6h. 45m., and faded away in the 

 moonlight between I5h. and i6h. astronomical time), and 

 whose beams converged several times from a large part of the 

 horizon towards the magnetic zenith (formation of corona was 

 noted at 7h. 2m., loh., and I3h.), we have had aurora on 

 February 14, 15, 24, 25, March I (at 7h. high arch, with the 

 highest part through a and ;8 Cephei, 7h. 55m. corona, between 

 8h. and loh. pulsating and flashing light, sometimes with 

 apparently screw-formed motion), March 2, 3, 6 (at loh. 

 curtains and corona, yellow-green colours), March 24, 25, 

 26, 27, April 23, 24 (at loh. lom. curtains, yellow-green) 

 April 25 (strong light visible through small openings in 

 cumulo-stratus in the north). The last display was on May I, 

 with corona at 9h. 40m., after loh. flashes, curtains, and 

 beams, at I3h. beams. About iih. there was a peculiar 

 downward motion of reddish light near the north horizon. 



The magnetic disturbances of February 13 were also the 

 greatest we have had for some years. The magnetometers, of 

 the Gaussian construction, are generally observed at 2h. and 

 2lh., but on February 13 observations were made every hour 

 from II p.m., in correspondence with Bosekop in Finmarken, 

 where the German observers, MM. Brendel and Baschin, were 

 taking magnetical observations and photographs of the aurora 

 during February and part of January. In Christiania the per- 

 turbations were comparatively small in declination (westerly 

 maximum 12° 35' noted at I2h. lOm., minimum li" 42' at 

 I5h. i8m., but neither of them absolute, the observations not 

 being continuous) ; but the horizontal intensity, which had 

 already begun to increase a little at 2ih., February 12, varied 

 by more than 003 C.G.S. units, a maximum of 0-171 having 

 been noted at 2h. 30m., and a minimum of about 0140 

 from I2h. om. to I3h. 20m. ; as the mirror of the magnet was 

 in both cases outside the scale, the values could only be roughly 

 measured. At i6h. om. the bifilar had returned to the 

 small end of the scale, but a nearly constant value of the 

 horizontal intensity was only attained after 5h., February 14. 

 The inclination had a maximum of 73° 18' at I3h. lom., from 

 which it gradually diminished, with some fluctuations, towards 

 the normal value, about 71° o'. 



With reference to Mr. Backhouse's observation of nacreous 



