348 



NA TURE 



{March 2, 1876 



With fuller information on the rare and unstable compounds, 

 than is to be found in ordinary text books (such as " Miller's 

 Elements"), no doubt numerous other instances might be 

 noticed ; these, however, will suffice to exemplify the general 

 law. 



With regard to black compounds, I have not included them 

 in the above list, as they only complicate it unnecessarily ; the 

 ordinary term black being used to describe a very minute amount 

 of any possible colour. It is only when the predominating 

 colour is observed (as blue- olive- red-black, &c. ), that the 

 description is of any value. " Brown " is likewise ill-defined in 

 its spectral position, as various tints called brown generally 

 include a small quantity of any colour except blue. Black and 

 brown compounds are therefore inconclusive without a spectral 

 examination of the colour in each case. 



The real anomalies to the above law are the following com- 

 pounds. CrClo white, CfjClg violet ; MnClj pale pink, MnCls 

 green brown; As^So red, As„S3 yellow; HgCl yellow white, 

 HgCla white ; Aul yellow, Aulj green. In these five pairs the 

 law is apparently reversed, but they cannot be said'to nullify an 

 induction from thirty-five pairs, as enumerated above. A con- 

 formity of six cases in seven to a rule is sufficient to establish it 

 as a law, from which the modifying causes have not yet been 

 eliminated. 



One or two other cases might at first seem to be also excep- 

 tions ; but as they are really salts, in which the electro-positive 

 of the base is the same element as that of the acid, they are not 

 necessarily to be compared with binary compounds. 



In the order of the colours, it will be observed that, in nume- 

 rous cases, white occurs between blue and yellow compounds ; 

 and there is only one instance of a violet compound with more 

 of the electro-negative element than its white connection. As 

 white light comes (in the natural arrangement) between blue 

 and yellow, this order is more in accordance with the spectrum 

 than is the order of metachromism announced by Mr. Ackroyd, 

 in which white occurs before violet. 



It may be worth notice that the electro-negative elements 

 (whose increase reddens the compounds) are, on the whole, 

 rather more red and yellow than the electro-positive, many of 

 which are bluish, and even dark blue, as Na. 



This order of colours in successive compounds of a series 

 (which might be called taxichromisaa) was observed by the writer 

 some few years ago, in connection with the spectral order in 

 metachromism ; the latter, however, was only traced through a 

 dozen or so, of the oxides, without pursuing the subject farther. 



Bromley, Kent W. M. Flinders Petrie 



Seasonal Order of Colour in Flowers 



I THINK it may be useful to mention, in reference to several 

 letters on this subject in Nature, that light appears to have no 

 direct influence on the tints of flowers. I quote the following 

 from Sachs's Textbook, Engl. transL, p. 675 : — 



' ' As long as sufficient quantities of assimilated material have 

 been previously accumulated, or are produced by green leaves 

 exposed to the light, flowers are developed even in continuous 

 deep darkness which are of normal size, form, and colour, with 

 perfect pollen and fertile ovules, ripening their fruits and pro- 

 ducing seeds capable of germination." 



W. T. Thiselton Dyer 



r Rainbow Projected on Blue Sky 



An instance of this phenomenon, which is referred to as rare 

 in Nature (vol. vii. p. 68), occurred to-day, Feb. 22, 4'30 p.m. 

 The sky was almost quite clear and a light shower of rain falling, 

 caused one to look upwards for the clouds whence it proceeded, 

 but the air was uniformly clear near the zenith, though bordering 

 the horizon all round there were some detached cumuli, and a 

 few thin filmy modificatioas higher in N.E. There were no 

 visible signs of the origin of the falling drops. On turning round 

 to east, a solar bow was seen, for the most part on a background 

 of azure. It was a complete bow and moderately intense. Near 

 the vertex it rested on thin clouds, as did the extremities on the 

 horizon, but they had no sensible effect on the phenomenon, for 

 it was observed that as they receded fiom the upper portion, the 

 bow remained intact and equally bright in all its parts. The 

 arms spanned areas of blue sky. It was very transient, like the 

 shower, and fading rapidly, was gone within 10 sec. after it was 

 first observed. At 4.45 P.M., the left side of an incomplete bow 

 was seen in N. E. on clear sky, except at the lower extremity : 



no rain accompanied it. It was as evanescent as the first. The 

 weather was very unsettled and showery with low barometer. 

 Thermometer 50°, with brisk wind from W.S.W. — 34 inches of 

 rain had fallen during the previous 9 days. 



Wii.i.iAM F. Dknning 

 Ashleydown, Bristol, Feb. 22 



OUR ASTRONOMICAL COLUMN 

 Olbers' Comet of 181 5. — The comet discovered by 

 Gibers, at Bremen, on March 6, 18 15, and with which his 

 name has been usually associated, belongs to the group, 

 the members of which revolve in periods a few years less 

 than the period of the planet Uranus. The deviation 

 from parabolic motion was remarked independently a 

 few months after the discovery by Bessel, Gauss, Olbers, 

 Nicolai, and Triesnecker. Bessel calculated elliptic 

 elements so early as the middle of May, and was followed 

 by Gauss in June, and the former subsequently investi- 

 gated, as completely as was practicable at the time, the 

 elements resulting from the whole course of observation 

 and the effect of perturbation in the actual revolution. 



When first detected, the comet is said to have been barely 

 visible in a good achromatic, and according to the Zeit- 

 schriftfiir Astronomie, was not discernible without tele- 

 scopic aid at any time ; it is, however, upon record that 

 in Russia it was seen with the naked eye. The last ob- 

 servation was made by Gauss at Gottingen, on August 25. 

 At discovery on March 6, at 10 P.M., it was in R.A. 49'''3 

 and N.P.D. 6i°*3, distant from the sun 1*47 and from the 

 earth 1*30, and at Gauss's final observation it was in 

 R.A. 2iy'^'i, N.P.D. 84°'5, distant from the sun 2'io, and 

 from the earth 2'36. Representing the intensity of light 

 at discovery (calculated according to the usual expression 



— — ~\ by tinity, a maximum of i*4 was attained on May 



3, and it had diminished 'on August 25 to o*i6, or one- 

 sixth of that at discovery. 



Bessel's determination of the elements of this comet 

 and of the perturbations to the next return will be found 

 in Abhandl. der BerL''Acad. Mathem. CL 1812-15. His 

 definitive figures are as follow : — 



Perihehon Passage, 1815, April 25-99867 M.T. at Paris. 



Longitude of the perihelion 149° i' 56" 



„ „ ascending node 83° 28' 34" 



Inclination to ecliptic 44° 29' 55" 



Excentricity 0"93I2I97 



Semi- axis major I7'63383 



These elements apply to the date of perihelion pas- 

 sage ; the longitudes from mean equinox of 18 15. 



To the above value of the semi- axis corresponds a 

 period of revolution of 74*049 years. The distance in 

 perihelion is i"2i3 (the earth's mean distance from the 

 sun being taken as ujiity), and the aphelion distance 

 34*055; the minor semi-axis is found, to be 6*427. The 

 ascending node is situate near the orbit of the planet 

 Mars, and if the actual form, of orbit is due to planetary 

 attraction, it is probably to be ascribed to a near approach 

 of the comet to Mars at some distant period. At the 

 other node the radius-vector is 3*81 in the region of the 

 minor planets. The comet, it will be seen, recedes be- 

 yond the orbit of Neptune, but near the aphelion it has 

 a depression of nearly 40° below the plane of the ecliptic. 



According to Bessel, the mean motion at the instant of 

 perihelion passage in 1815 corresponded to a period of 

 revolution of 27046"9 days, and he found that this would 

 be diminished 824*5 ^^y^ ^Y ^^^ united action of the 

 planets Jupiter, Saturn, and Uranus, thus obtaining 

 26222'4 days for the actual revolution, and fixing the 

 next arrival at perihelion to 1887, February 9*4, Beyond 

 doubt, however, this date may now admit of a closer 

 determination, and very probably we may soon hear of a 

 further investigation being undertaken. The original obser- 

 vations made at the Observatory of Paris, and those at one 

 or two other observatories in greater or less detail, are 



