=

sistance, or the liquids. Now it is evident, from the fact of the copper wire being heated much less than the platinum, and increasing its resistance in a higher degree than the latter (E. Becquerel gives the relations between resistance and temperature, for platinum =100+0.1861 t.° Cels., for copper 100+04097 t.° Celsius), that the ratio of increase of resistance of the copper wire should appear to be remarkably smaller than that of the liquids which constitute the main part of the internal resistance, under equal cicumstances. The explanation above referred to, so acceptable in case of the internal resistance of the liquids, seemed therefore to be a failure in case of the copper wire, since I presumed my experiments accurate enough at least to trace the difference, naturally to be expected, between the ratio of increase of resistance of liquids and that of copper.

But there is a point of far greater weight to be considered. The common rule for the calculation of the constants of a battery, by combining the highest direct intensity with some lower ones, or any other specific "method" certain experimenters prefer, is rather arbitrary, since Ohm's formula knows of no such restrictions, but permits to combine any two or more observations. And in thus combining any two observations I at first expected to get more reliable results, especially by throwing out those observations in which the platinum wire still had been heated very much. But these calculations seemed to run perfectly wild, giving much greater values than ever before. Combining, for instance, the intensity 0-115 with 100 centimeters of platinum wire, in table I, with each of the following three intensities, after the formula RP2P, the values for the in

=

ternal resistance became respectively 23-75, 31-67, 21:15, with a mean of 25:52, which, compared with the first value for the resistance (3:48), gave an increase from 1:7-33, thus far greater than anybody could attempt to explain by way of influence of temperature. I now became pretty much convinced that there exists indeed some other reason for the increase of resistance with decrease of intensity, and a reason very much more powerful than the influence of temperature, and one able to conceal, by way of its great ratio and the errors of observations, comparatively slight differences in the ratio of increase of resistance of liquids and copper as proceeding from the influence of temper

ature.

On the other hand, those great and much varying values secmed to pronounce my observations as perfectly worthless. could not imagine that facts of such important bearing could have been overlooked. Though, it is true, the experimental results the best observers have arrived at do not harmonize with

each other to such a degree as to silence every doubt about their reliability.

However, my experiments were open to grave objections, as visibly indicated by the irregularity of the results. Indeed, the use of acids (nitric acid particularly) from former experiments may have caused some polarization to set in. The time required for a series of observations-the diminishing of the direct intensity usually occurring during this time-the change of the temperature and chemical composition of the liquids-the influence perhaps existing of the circuit being kept closed during the rapid but gradual change from one external resistance to another-all these circumstances may be considered susceptible of bringing on such extraordinary results as those just referred to.

TABLE I.

Bunsen's battery. Gas coke in nitric acid. The acids had been used previous to the experiments. The platinum wire, when shorter than 8 centimeters, became red hot.

The same battery. The gas coke, after washing with water, had been exposed to the drying action of the air for 24 hours. The same acids.

The same battery, except the gas coke exchanged for another dry piece not used before. Acids the same.

TABLE IV.-Bunsen's battery, with a common hollow cylinder. The acids had been used previously.

TABLE V.-Bunsen's battery, with gas coke. The acids had been used previously. The nitric acid, however, was mixed with sulphuric acid to restore the strength, The circuit contained 80 inches of thin copper wire.

0

....

Rheo- Com. Tan- Int. of 80 in. inter. Rheo Com- Tan-
chord. pass. gent. resist. copper. resist. chord.
34.5°
.6964

pass. gent.

Int. of 80 in. inter. resist. copper. resist.

40

9-7

•1709

50

8.5

1495

8.34 377 4:57

100

9.57 4:32 5.25

120

12 17.4

⚫3134 9.82 4.44 5.38

The 80 inches of copper wire excluded, and the circuit closed directly, the cur51.20 52.4

rent gave

at the compass, corresponding to a mean tangent of 1-2712.

(To be continued.)

.1

ART. LII.-On the Spectrum of a new Star in Corona Borealis ;' by WILLIAM HUGGINS, F.R.S., and W. A. MILLER, M.D., Treas. R.S.'

YESTERDAY, May the 16th, one of us received a note from Mr. John Birmingham of Tuam, stating that he had observed on the night of May 12 a new star in the constellation of Corona

The Astronomer Royal wrote to one of us on the 18th, "Last night we got a meridian observation of it; on a rough reduction its elements are

agreeing precisely with Argelander, No. 2765 of 'Bonner Sternverzeichniss,' declination +26°, magnitude 9.5." Mr. Baxendell writes on the 21st, "It is probable that this star will turn out to be a variable of long or irregular period, and it may be conveniently at once designated T Corona." Sir John Herschel informs one of us that on June 9, 1842, he saw a star of the sixth magnitude in Corona very nearly in the place of this strange star. As Sir John Herschel's position was laid down merely by naked eye allineations, the star seen by him may have been possibly a former temporary outburst of light in this remarkable object.

* From the Proceedings of the Royal Society, No. 84, 1866.

AM. JOUR. SCI.-SECOND SERIES, VOL. XLII, No. 126.-Nov., 1866.

Borealis. He describes the star as "very brilliant, of about the 2d magnitude." Also Mr. Baxendell of Manchester wrote to one of us, giving the observations which follow of the new star, as seen by him on the night of the 15th instant:

"A new star has suddenly burst forth in Corona. It is somewhat less than a degree distant from & of that constellation in a southeasterly direction, and last night was fully equal in brilliancy to Serpentis or Herculis, both stars of about the 3d magnitude."

Last night, May 16th, we observed this remarkable object. The star appeared to us considerably below the 3d magnitude, but brighter than & Coronæ. In the telescope it was surrounded with a faint nebulous haze, extending to a considerable distance, and gradually fading away at the boundary. A comparative examination of neighboring stars showed that this nebulosity really existed about the star. When the spectroscope was placed on the telescope, the light of this new star formed a spectrum unlike that of any celestial body which we have hitherto examined. The light of the star is compound, and has emanated from two different sources. Each light forms its own spectrum. In the instrument these spectra appear superposed. The principal spectrum is analogous to that of the sun, and is evidently formed by the light of an incandescent solid or liquid photosphere, which has suffered absorption by the vapors of an envelope cooler than itself. The second spectrum consists of a few bright lines, which indicate that the light by which it is formed was emitted by matter in the state of luminous gas." These spectra are represented with considerable approximative accuracy in the annexed diagram.

Spectrum of absorption and spectrum of bright lines forming the compound spectrum of a new star near & Corona Borealis.

On the 17th this nebulosity was suspected only; on the 19th and 21st it was

not seen.

The position of the groups of dark lines shows that the light of the photosphere, after passing through the absorbent atmosphere, is yellow. The light, however, of the green and blue bright lines makes up to some extent for the green and blue rays (of other refrangibilities) which have been stopped by absorption. To the eye, therefore, the star appears nearly white. However, as the star flickers, there may be noticed an occasional preponderance of yellow or blue. Mr. Baxendell, without knowing the results of prismatic analysis, describes the impression he received to be "as if the yellow of the star were seen through an overlying film of

a blue tint."

Description of the spectrum of absorption. In the red a little more refrangible than Fraunhofer's Care two strong dark lines. The interval between these and a line a little less refrangible than D is shaded by a number of fine lines very near each other. A less strongly marked line is seen about the place of solar D. Between D and a portion of the spectrum about the place of b of the solar spectrum, the lines of absorption are nuinerous, but very thin and faint. A little beyond b commences a series of close groups of strong lines; these follow each other at small intervals, as far as the spectrum can be traced.

Description of the gaseous spectrum.-A bright line, much more brilliant than the part of the continuous spectrum upon which it falls, occupies a position which several measures make to be coincident with Fraunhofer's F. At rather more than one-fourth of the distance which separates F and G, a second and less brilliant line was seen. Both these lines were narrow and sharply defined. Beyond these lines, and at a distance a little more than one-third of that which separates the second bright line from the strongest bright one, a third bright line was observed. The appearance of this line suggested that it was either double or undefined at the edges. In the more refrangible part of the spectrum, probably not far from G of the solar spectrum, glimpses were obtained of a fourth and a faint bright line. At the extreme end of the visible part of the less refrangible end of the spectrum, about C, appeared a line brighter than the normal relative brilliancy of this part of the spectrum. The brightness of this line, however, was not nearly so marked in proportion to that of the part of the spectrum where it occurs, as was that of the lines in the green and blue."

General conclusions. It is difficult to imagine the present physical constitution of this remarkable object. There must be a photosphere of matter in the solid or liquid state emitting light of all refrangibilities. Surrounding this must exist also an at

On the 17th, the lines of hydrogen, produced by taking the induction-spark through the vapor of water, were compared in the instrument simultaneously with the bright lines of the star. The brightest line coincided with the middle of the expanded line of hydrogen which corresponds to Fraunhofer's F. On account of the faintness of the red end of the spectrum, when the amount of dispersion necessary for these observations was employed, the exact coincidence of the line in this part of the spectrum with the red line of hydrogen, though extremely probable, was not determined with equal certainty.

The spectra of the star were observed again on the 17th, the 19th, the 21st, and the 23d. On these evenings no important alteration had taken place. On the 17th and succeeding evenings, though the spectrum of the waning star was fainter than on the 16th, the red bright line appeared a little brighter relatively to the green and blue bright lines. On the 19th and 21st the absorption lines about b were stronger than on the 16th. From the 16th the continuous spectrum dimin ished in brightness more rapidly than the gaseous spectrum, so than on the 23d, though the spectrum as a whole was faint, the bright lines were brilliant when compared with the continuous spectrum.