38 



NATURE 



{Jtine 6, 1889 



"lutely regular, as in a homogeneous gas, for the reason that the 

 meteorites are not equally distributed. 



That there is such a shortening was proved by Encke for the 

 •comet which bears his name, as the folio a ing table will show : — 



Returns of Enckes Comet, showing Kedttced Period of 

 Revohttion. 



Observed Period t-v-a- 



of Revolution. Difference. 



day.s. h. m. h. m. 



J^rom 1786 to 1795 three times 1212 15 7 



,, 1795 ,, 1805 ,, ,, 1212 12 o ^ ' 



,, 1805 ,, 1819 four ,, 1212 o 2g ^J. 3' 



,, 1819 ,, 1822 1211 15 50 3^ 



,, 1822 ,, 1825 1211 13 12 : ^ ^o 



,, 1825 „ 1829 1211 10 34 ^ ft 



,, 1829 ,, 1832 1211 741 ^ ^^ 



» 1832 „ 1835 1211 5 17 ^ ;4 



„ 1835 ,, 1838 1211 238 ^39 



,, 1838 ,, 1842 1210 23 31 ^ ^ 



,, 1842 ,, 1845 i2ro 21 7 '^ -% 



„ l84q ,, 1848 I2IO 18 29 ^ 3» 



1848 ,, 1852 .* I2IO 17 2 I ^l 



,, 1852 ,, 1855 I2IO II 17 i 5 ^l 



„ 1855 „ 1858 1209 13 41 I ^^ 3& 



Here, then, we have three possible sources of collisions. In 

 any case, if any light be produced by collisions, we have the 

 spectroscope as a sure guide to enable us to determine its 

 chemical origin. 



We have already seen that the telescopic appearance of a 

 comet when far away from the sun and when close to it are very 

 different. We must now introduce the verdict of the spectroscope. 

 It was observed by Dr. Muggins in the comets of 1866 and 1867 

 that when they were very far away from the sun the spectrum con- 

 sisted chieflf of a line seen in the spectrum of those nebul-jc which 

 he had up to that time examined. Unless, then, Dr. Huggins 

 has withdrawn this observation, there is a sfeetroscofic connection 

 between nebulae and comets away fnm the sun. 



The phenomena of comets revealed by the telescope show, 

 tis we have seen, that as a matter of fact a good many of them 

 seem to be connected in some way or other with the production 

 of luminous concentric or eccentric envelopes. 



In the case of a comet gradually getting nearer the sun, and 

 getting very excited as it gets there, we pass from the spectrum 

 already described to a very different one. There is a consider- 

 able change similar to that observed in experiments with 

 meteorites, the spectrum of carbon produced from some com- i 

 pound of carbon or another. In nineteen cases out of twenty ! 

 when the comet gets near the sun and near enough to the earth 

 for us to have a good look at it, the spectrum is a spectrum of 

 carbon. 



On its first appearance in a cometary spectrum, carbon is re- 

 presented by the flutings which are special to low temperatures. 

 In the most visible' part of the spectrum these flutings differ very 

 little in position from those which appear at a higher tempera- 

 ture, but in the blue there is a low-temperature fluting about 

 wave length 483, whilst the nearest high-temperature fluting is 

 474. If, therefore, this fluting be observed, the presence of cool 

 ■carbon may be safely inferred, although it would not be quite 

 safe to infer its presence from observations of the green flutings. 

 This has certainly been observed in two comets — namely, 

 Winnecke's comet (1868) on June 17 (the perihelion passage 

 occurring on June 24), and in Brorsen's cDmet (1879) on March 

 25 (the perihelion passage occurring on March 30). The limited 

 number of recorded appearances of cool carbon in comets is doubt- 

 less due to the same cause as in the case of the line near A 500, 

 which Dr. Huggins ascribed to an unknown form of nitrogen, 

 while I ascribe it to magnesium, since we know that there is 

 magnesium in meteorites, and we do not know that there is an 

 unknown form of nitrogen. The reason is that the temperature 

 being low, the light is excessively feeble and observations therefore I 

 difficult. When nearer perihelion passage, the comets get hotter, 

 and the spectrum of cool carbon is replaced by that of hot carbon. 

 Under these conditions of increased temperature, comets lend 

 themselves best to spectroscopic study, and hence it happens 

 that in the majority of cases the spectrum of a- comet (if the 

 ternperature be increasing) has not been observed until it has 

 ^arrived at this stage. 



Manganese is the next substance which writes its record in the 



spectroscope. It is first represented by a fluting at 558,^ which is 

 the brightest fluting in its spectrum at a low temperature. 

 This fluting is very persistent, and becomes visible even when 

 there is only a very small percentage of manganese present in 

 the substance examined. The fluting is always seen before the 

 iron lines in the spectrum of ordinary iron at the temperature of 

 the oxyhydrogen flame, and this is the case even with the purest 

 specimens of electrolytic iron which has yet been prepared. The 

 effect of the addition of this fluting to the spectrum of carbon 

 is to modify the appearance of the citron band in the cometary 

 ' spectrum in a very definite manner. 



At a still higher temperature, the radiation of lead is added 

 to that of manganese and carbon, which still further modifies the 

 appearance of the citron band. The brightest lead fluting is at 

 wave-length 546, and when this is present in the spectrum of a 

 comet the citron band has three maxima of brightness, one at 

 564 (carbon), one at 558 (manganese), and one at 546 (lead). 



Afterwards, the temperature having increased, the radiations of 

 manganese and lead give way to the absorption flutings of thest- 

 substances, carbon radiation from the interspaces still remaining. 

 The result is again a very definite modification of the appear- 

 ance of the citron cometary band, the general effect being an 

 apparent shifting of the carbon fluting from wave-length 564 to a 

 more refrangible part of the spectrum — namely, to 558, when only 

 manganese absorption is added, and to 546 when both manganese 

 and lead absorptions are added. 



Until quite recently, the variations in the position of the citron 

 band in different comet=, or in the same comet at different 

 periods, have been attributed to faulty observations, it being 

 supposed that carbon pure and simple was in question. It i 

 now certain, however, that this is not so in all cases. Tli 

 variations are real, and are simply dependent upon the tempera- 

 ture, or indirectly upon the distance from perihelion. 



In some cases iron fluting absorption has also been observed 

 under these conditions of high temperature. 



The spectral conditions brought about in the comets which 

 in our time have got nearest to the sun were \ciy siinilar to 

 those observed in the electric arc, and the rccoided obser- 

 vations of the spectrum show that we were dealing with a 

 considerable nu nber of lines of iron, manganese, and other 

 substances. 



We see in the telescope that a comet puts on the appearance 

 of a centra] nucleus with surrounding envelopes or jets, so that 

 we must understand that in the spectroscipe the spectrum ol 

 the nucleus is seen distinct from the spectrum of the envelope^ 

 and jets, because the former is made to fall upon one part of the 

 slit of the spectroscope and the latter upon another. 



When a comet a}jproaches very near to the sun we get in 

 addition to the usual flutings of carbon, bright lines, especially in 

 the spxlrnin of the mcclacs, so that in addition to the long fluting> 

 of carbon as visible in the spectroscope we have short lines 

 added along the nucleus in the red, yellow, green, and so on. 



In those comets which have reached a very high temperature, 

 like Comet Wells and the Great Comet of 1882, there is 

 evidence of line absorption. At the same time there were 

 bright lines, proceeding from the incandescent vapours driven 

 away from the meteoritic nuclei by the solar repulsion. Without 

 this repulsion, it is highly probable that there would be line 

 absorption pure and simple, and this has to be taken into 

 account in comparing the spectra of comets with the spectra of 

 other meteor-swarms. 



Daring the passage of a comet from perihelion to aphelion, 

 the temperature decreasing, these changes take place in inverse 

 order. 



In spectral phenomena, then, we have another term of com- 

 parison to apply, and it may be stated that the sequence of 

 spe :tral changes are now known to us in a very definite way, so 

 that the chemical changes which take place in the composition of 

 the vapours of comets produced by collisions at various distances 

 from the sun have been ascertained. 



Whether we take form, distance apart of component parts, or 

 spectrum, there is now ample proof that the external bodie> 

 which supply us with these shreds and patches which we term 

 comets are the nebuhe. 



It has already been stated that if we can rely upon Dr. Muggins's 

 observations, in some comets at aphelion and in some planetary 

 nebulae we get a single line at the same wave-length, so that from 



I Students of spectrum analysis will understand that thi5 is a "short 

 title," and does not represent the exacz wave-length, which with adequate 

 instruments might require something between loand 100 numerals 



J 



