4 EXPKIUMENTS WITH lONlZKU AIK. 



"dust" contents of air entering C, with the aid of a "duster" at h, and a variety of 

 similar j>ur[)oses. It is to be noted tliat the tubes E, D, E, must all pass out of the 

 room. If, for instance, b opened into the room (often very desirable), then if the 

 jet is only slightly in action or not at all, cold air will pass into C and I) and out 

 at />, {IS well as at Z/, into the room. The flow of dusty air would therefore be irreg- 

 ular. An advantasre is secured in makint' the common tube C^lonEr, so that the air 

 may be well mi.xed before impiuiring upon the jet. At best, however, air dusted in 

 this way is but an inferior substitute for good room air, and the results show more 

 fluftuation. 



2. Jet. — A jet of almost any kind, with a smooth round hole about 1 milli- 

 meter, is satisfactory. It may be conveniently sha|)ed fiom lead pipe. \Mien the 

 variations of color phenomena are to be quantitatively studied it is desirable to make 

 the jet as shown in figure 6. Here ab is a brass tube about 10 cm. long and 1 cm. 

 in diameter, the end b of which is threaded both on the outside and inside, so that 

 a thin-walled nipple, e, closed at the outer end, may be in.serted. Into the top of 

 c a hole, d, is smoothly drilled. The outer thread at b is useful in screwing the 

 jet to the color tube, in to the centre of which c projects. The open end a of the 

 jet is connected with the tube b of the steam box, figure 5, In a union. 



The jet may either be fed with pure steam or with compressed air super- 

 saturated with aqueous vajior. When steam is used alone, it must be available in 

 lai-ge (piantity at a pressure up to one atmosi)here or more. This pressui-e may 

 then be reduced by an ordinary steam-cock. To dry the steam it is convenientlv 

 passed into a cylindrical box, A, figure 5, called a separatoi-, through the vertical 

 pipe a, surmounted by the cock specified. It is conveyed to the jet by a lateral 

 pipe, b. The water which collects is discharged through c, and a pipe, o?, communi- 

 cates with an open mercury mouometer for the measurement of pressure. The 

 separator shoidd be made of gas-fittei-s' appurtenances and suitaldy jacketed. 



3. Available color sequences. — The next question to be preliminarily disposed 

 of is the succession of colors. I will arrange them in table 1, statins; biiefly how 

 they were obtained and placing the colors where I think they belong. I will also 

 add Quincke's' revision of the colors of Newton's rings, as quoted by Kohlrausch,- 

 for comparison. Kiessling's' results need some explanation. His first series of 

 colors is produced by compressing surcharged, suitably dusted air, and then allow- 

 ing it to expand suddenly to atmospheric pressure. The experiment being made 

 in a glass sphere illuminated by sunlight, the colors are observed in the axis of 

 illumination. The succeeding series are obtained by aid of an exhaust pump. 

 They are less easily located relatively to Newton's scale (see last columns) than 

 Kiessling's first series, which is intensely brilliant. Aitken,'* who exhausts Ion"' 

 tubes, gets a wider range of colors, running, as I interpreted them, from the first 



' Quincke: Po^g. Ann., CX.XIX, p. i8o, i866. 

 ' Kohlrausch : Leitfaden, sth edition, p. 340. 



' Kiessling : Ddmmerungsfrscheiniingcn, p. 140, 1888 ; Hamburg, L. Voss. 



* Aitken : Trans. Roy. Soc. Edin., XXXV, part I, 1S88 ; ibid., XXXVII, part J, No. 3, 1892 ; 

 Proc. Roy. Soc. Edin., XVI, p. 135, 1889 (containing full descripiions); ibid., XVIIli p. 25J, 1S90-' 



