January 7, 1887.] 



SCIEJSrCE. 



13 



discharge current due to the first charging. An ex- 

 perimental coil was then made up of 1,800 ohms of 

 wire having unparaffined cotton insulation. It was 

 wound on a warm rainy day, and tested immediately, 

 showing the strongest polarization found, driving the 

 spot of light violently off the scale. The coil was 

 then baked in a hot-air oven at 150° C. for an hour, 

 and tested again when cool. No trace of polariza- 

 tion coiild then be fovind, though the charging cur- 

 rent was increased. The previous observations of 

 course indicated electrolytic polarization as the dis- 

 turbing cause ; and the last showed, that, in the case 

 of that coil, it was electrolysfs of water absorbed 

 from the air by the cotton insulation. The experi- 

 mental coil was then heated, and soaked well with 

 pure iDaraffine, and drained while hot until it seemed 

 to be as nearly as possible in the same condition as 

 the 40,000 Elliott coil, and tested when cool. No 

 trace of polarization was shown. It was then put 

 aside in the instrument case to see whether it could 

 still absorb water enough to polarize. Ten days 

 later, just after the Buffalo meeting, the coil was 

 tested again and polarized strongly. On heating it 

 again, the polarization entirely disappeared. A drop 

 of hydrant water placed on the coil caused polariza- 

 tion to re-appear in five seconds, and in five minutes 

 the effect was so strong as to drive the needle to its 

 stops. 



The degree of error in measurement resulting from 

 polarization was not examined, but Professor Men- 

 denhall's statements show that it may be a consider- 

 able quantity. 



It is obvious that unparaffined coils are. on this ac- 

 count, unsuited to the best work ; also that coils well 

 paraffined (as in the B.A. unit coil) or coils freshly 

 baked and paraffined are free from such error. 



The paraffining of ordinary coils, even when as 

 thoroughly done as by the Elliotts, is not a perma- 

 nent protection, probably because of cracking of the 

 mass of paraffine, allowing vapor to reach the wire 

 and insulation. A test will quickly determine the 

 condition of any partictilar coil. A box might be 

 made proof against polarization by filling entirely 

 the space about the freshly baked coils with pure 

 paraffine, just warm enough to flow freely. Tempera- 

 ture difficulties could be in part overcome by thermo- 

 junctions, as in standards. Another and on some 

 accounts better plan would be to mount the coils in 

 an impervious box with liquid-tight joints, and fill- 

 ing the interior with a petroleum oil, which may 

 readily be found in market, of such quality as to 

 exhibit no polarization. With such a box, there 

 need be no uncertainty as to the temperature of the 

 coils. Benj. F. Thomas. 



Columbus, O., Dec. 27. 



Atmospheric lines in the solar spectrum. 



The ingenious device recently published by Mr. 

 Conner, for detecting the lines in the solar spectrum 

 due to the earth's atmosphere, recalls a similar plan 

 proposed by the writer some years ago. In a letter 

 dated Feb. 21, 1883, I wrote to Professor Kowland, 

 " I hope that you will try the experiment of which 

 I spoke to you last summer, — forming two images of 

 the sun, and photographing the spectra of the oppo- 

 site limbs. A glance would serve to distinguish the 

 solar from the telluric lines." An accompanying 

 sketch showed that a double-image prism was to be 

 placed between the slit and a lens forming an image 



of the sun upon it.- This prism was to be moved 

 until the two images were in contact. The east and 

 west limbs were thus brought together, and the slit 

 was placed at right angles to their line of junction. 

 In the photograph, telluric lines should cross the 

 spectrum undeviated, while solar lines would be bent 

 in opposite directions where they crossed the line of 

 separation of the two spectra. The advantages of 

 this method over that of Mr. Conner are, first, its 

 simplicity, as it is easily tried by any one who has 

 a spectroscope giving a sufficient diffusion ; secondly, 

 the solar lines, instead of becoming hazy, continue 

 well defined. For these reasons I call attention to 

 the matter, and not to detract from the credit due to 

 the eminent French physicist, who has preceded me 

 both in trying and publishing a solution of this very 

 important problem. Edward C. Pickeeing. 



Harvard coll. observ., Jan. 1, 1887. 



A brilliant meteor. 



On Jan. 3, 1887, at 5.15 p.m., I observed a meteor 

 of unusual brilliancy. It started, as nearly as I could 

 make out, from the constellation Ursa Minor, pos- 

 sibly a little higher up, moving with a rapid rush 

 and brilliant light in an easterly direction. As it 

 neared the horizon, its speed apparently diminished, 

 until it disappeared behind some trees. It was visi- 

 ble fully thirty seconds, and, during the last part of 

 its flight, appeared to float slowly downwards. A 

 trail of considerable length was drawn behind, giving 

 it the appearance of a large rocket. Its flight was 

 unattended by any sound. E. W. Wood, Jr. 



Jamaica Plain, Mass. 



What was the rose of Sharon ? 



I notice in your issue of Dec. 31 an article on the 

 rose of Sharon. Without desiring to enter into the 

 discussion of this svxbject, I wish to refer those in- 

 terested to a few words upon this subject by an emi- 

 nent investigator. Speaking of that part of the 

 pleistocene plain near Jaffa, bordering the Mediter- 

 ranean Sea, Sir J. W. Dawson, in his recent work on 

 'Egypt and Syria,' says, " In February we found it 

 gay with the beautiful crimson anemone (A. coro- 

 naria), which we were quite willing to accept as the 

 ' rose of Sharon,' while a little yellowish-white iris, 

 of more modest appearance, growing along with it, 

 represented the ' lily-of-the-valley ' of Solomon's 

 song." From this would it not be reasonable to in- 

 fer that this anemone is quite generally recognized 

 as the ' rose of Sharon ' ? Amos W. Butlee. 



Brookville, Ind., Jan. 3, 1887. 



Electrical phenomena on a mountain. 



In confirmation of the observations of M. F. 

 {Science, viii. p. 564) in relation to electrical jphe- 

 nomena on Lone Mountain, near Bozeman, I beg leave 

 to call attention to the fact that more than twelve 

 years ago Mr. Franklin Ehoda, assistant topographer, 

 in his ' Eeport on the toj)ography of the San Juan 

 country ' {vide F. V. Hayden's Report of U. S. geolo- 

 gical and geographical survey of the territories for 

 the year 1874, pp. 456-458, also p. 461), gives a 

 detailed and graj)hic account of similar electrical 

 manifestations experienced by Mr. A. D. Wilson and 



