24 THE STRUCTUKE OF THE ITUOLEUB. 



rrimiiiation. Tbe charge of nuclei is removed too rapidly by tlie condensation, the 

 fall ])einLr too small to admit of much evaporatitm. If the tube is horizontal, two 

 longitudinal vortices, its shown in figure la, are seen immediately after exhaustion, 

 as the air heated by the sides ()f the tube rises on the outside and descends into 

 the axis. Fog is soon dissipated in this way (a condition of things which may be 

 traced even in very wide vessels), oi- else its homogeneity is imi)aired. When the 

 tube is hehl vertical, there is a corresponding difficulty. Under no circumstances 

 is there a guaranty that the charge of nuclei added is ever uniforndy distributed. 

 It will be of unequal density along the length. 



Tubes have an advantage inasmuch as they insure the occuri-ence of truly 

 axial colors, and for this reason I spent some time with them. The general form 

 of tube is given in figure 1, AA being the tube, d the tubulure for exhaustion, c 

 the tubulure for influx of pure air through the filter, /, containing compressed 

 cotton. When the cotton is removed, nuclei may be admitted here. AVhite lisrht, 

 L, enters Ijy the window, b, and is observed through a. 



Even with tubes but 3 cm. in diameter and less than a meter long, the axial 

 colors are strong, but the dull colors soon vanish. Immediately after chai'ging 

 with nuclei, dense opaques are the result of the smallest exhaustions. No special 

 advantage was gained favorable to purity of color when the diameters of the metal 

 tubes were increased to 5 cm., the tubes being 2 meters long. In case of glass 

 tubes of these dimensions illuminated by sunlight, the gi'adual extinction of the 

 li^-ht alon" the axis is well observed. The diflfraction colors are also seen on 

 oblique vision from the outside of the tube. Tiie end colors are of no value and 

 the light scarcely penetiates beyond one meter. 



4. Vise. — Incidentally I also tried a disc-shaped apparatus, figure 2, consisting 

 of two glass plates, _y and i/, secured by a metallic ring, /•. "\Vitlio\it i-emoving the 

 above difficulties, the colors even of the coronas were here very weak and the 

 apparatus too fragile. The disc is a close ajiproach to a surface apparatus, how- 

 ever, and from this i>oint of vie\v, often useful. 



5. (rlohes. — For studying coi'onal colors, spherical receivers. A, figure 3, as 

 used by C()ulier and by Kiessling. are preferable. Their diameter should exceed 



25 cm. With divergent sunlight the display is gorgeous, the colors glowing 

 metallically. The mantel of a Welsbach l>urner W, seen through a round hole, a, 

 in the screen, s, is better for the usual observational purposes. To an eye at JS, 

 the axial colors (colors seen in the color tube' in case of the steam jet) are not 

 very vivid, for the al)sorption along the column ab is not sufficient to subdue the 

 intensity of white light from II'. The diffraction colors, as acb, are splendid, par- 

 ticularly on blotting out the axial beam, ab, liy a small black circular screen, pasted 

 to the sphere in front of the eye. The vortices suggested in figure la, are now no 

 longer violent. They serve rather a useful pur'pose in keeping the contents of the 

 receiver homogeneous as to dust contents. It is with this apparatus that the 

 colored central ]iatches of the coronas, carefully to be distinguished from the axial 

 colors, are best observed. The lid of the globe secures the gauge G, the exhaust 

 pipe d and the filter /', all with approjiriate stop-cocks. 



' Experiments willi loin':, J Air 1. c , (^haj). I. 



J 



