474 



NA TURE 



IMarch 18, 1886 



tion was to find the tlifference between these, with respect to 

 the intensities of their total and residual magnetisation, due to 

 different degrees of magnetising force. 



The apparatus was ananged, and tlie experiments made, 

 according to a simple magnetomelric method fully detailed in 

 the paper. The magnetising currents were derived from a bat- 

 tery of Thomson's tray Daniells, so arranged that any number 

 of cells could readily be placed in the circuit. 



The results represent the effect of a current gradually in- 

 creased from o to the maximum, gradually diminished to o again, 

 and of the same process repeated with a negative current. They 

 are shown in curves, the abscis^^e of which are proportional to 

 the magnetising forces, and the ordinates to the magnetisation 

 produced. Figures are given by which to reduce these values 

 to absolute measure. 



It has been found that the " charcoal iron" has the highest 

 magnetisability, and the "soft steel" the lowest ; while that of 

 the "soft Scotch iron " approaches the former. With regard 

 to retentiveness, the " charcoal iron" shows the least, and tlie 

 "soft steel" the greatest. Annealing the latter, however, has 

 the effect of bringing it very near the "common wire," both in 

 respect of magnetisability and retentiveness. The two speci- 

 mens of cast iron differ considerably. The malleable iron 

 bar shows a very much higher magnetisability than the cast iron 

 ones, and its residual magnetisation was so low that it could 

 not be observed with the same arrangement of apparatus. 



These curves also present certain anomalies which are worth 

 investigation. The space about the zero on an enlarged scale is 

 affixed to each set to show its peculiarities more clearly. 



in the curves representing the residual magnetisation we find 

 a loop between the direct and return curves, more or less marked 

 in all the diagrams. A similar feature presents itself in the 

 curves of total mngnetisation in two of the diagrams, and there 

 seems to be a tendency always to form this loop. 



In that part of the return curve which represents the effects 

 of the .-mall magnetising forces, the residual magnetism is .'een 

 first to take a greater value, and then to diminish again just 

 before the zero of magnetising force is reached. This may be 

 observed also in the negative return curve. 



January 21. — " On Radiant Matter Spectroscopy : Note on 

 the Spectra of Erbia." By William Crookes, F. R. S. 



I have recently succeeded in getting the earth erbia in a 

 sufficiently pure state to allow me to examine its phosphorescent 

 spectrum without the interference which might be produced by 

 the presence of yttria, samaria, holmia, thulia, Ya, or ytterbia. 

 As in the case of yttria, ' the spectrum is best seen when erbic 

 sulphate is heated to redness and .submitted to the electric dis- 

 charge in a high vacuum. The addition of calcic sulphate inter- 

 feres with the purity of the spectrum. In this respect erbia 

 differs from samaria, as the latter earth seea.s to require the 

 presence of some other metal to develop its phophorescent 

 properties. 



The phosphorescent spectrum of erbia consists of four green 

 bands, of which the following measurements have been taken : — 



Scale of I 



speclrosci pe a p Remarks 



975° ■■• 5564 .. 3230 ... Approximate centre of a 

 wide band, shading off 

 at each side. 

 9'65o ... 5450 ... 3367 ... Approximate centre of a 

 band, narrower and 

 somewhat fainter than 

 the first hand. 

 9'525 ■■■ 531S ... 3536 ... Approximate centre of a 

 narrow band, bright and 

 moderately sharp on 

 each side. 

 9'40o ... 5197 ... 3702 ... Approximate centre of a 

 band, similar in appear- 

 ance to the first band, 

 but brighter. 

 These bands do not correspond in position to any in either 

 the yttrium or samarium .spectrum. The nearest approach to a 

 coincidence is between the first erbia green and the samarium 

 green, but when the two spectra are examined one over the 

 other it is seen that the samarium band is less refrangible than 

 the erbium band. 



^ Phii. Trans , part iii., 18S3, p. 913 (par. 71). 



The first green of Ya occurs midway between the first and 

 second greens of erbia, and the second Ya green comes between 

 the second and third erbia greens. 



Pure erbia is of a beautiful rose-pink colour. When illumin- 

 ated by sun or electric light and examined in the spectroscope it 

 gives a spectrum of black lines and bands as sharp and distinct 

 as the Fraunhofer lines. It is strange that this most character- 

 istic property has been recorded by so few observers. Indeed, 

 the only notice of it I have come across is a passing remark of 

 Prof. Cleve's that "the light reflected by dry erbia shows 

 absorption bands." 



The absorption spectrum given by a solution of pure erbic 

 chloride differs in some respects from the drawings mapped from 

 older observations, as the absorption lines of holmia and thulia 

 are absent. The fine group of lines in the green of the reflec- 

 tion spectrum is also absent in the absorption spectrum. 



The spectrum of bright lines emitted when erbia is rendered 

 incandescent in the blow-pipe flame has been often observed, 

 but the lines in this case are luminous on a fainter continuous 

 background, and are not particularly sharp, whilst the reflection 

 spectrum consists of black lines sharply defined on a continuous 

 spectrum. 



February 4. — " A Further Inquiry into a Special Colour- 

 Relation between the Larva of Smeri7i:hiis ocellatus and its 

 Food-plants." By Edward B. Poulton, M.A,, of Jesus and 

 Keble Colleges, Oxford. Communicated by Prof. J. S. Burdon- 

 Sanderson, F. K . S. 



The object of the paper was to give an account of the investi- 

 gation upon this subject which had been undertaken in 1885, 

 having been continued from the previous year (described in 

 Proc. Koy. Soc, No. 237, 1885, p. 269) The points which 

 had been raised, and upon which evidence was desired, were as 

 follows : — (i) The larvie are generally uniformly coloured on the 

 same food-plant, but someumes there are exceptions; can any 

 of these be due to the hereditary transmission of the influence of 

 food-plants upon the parent larvK ? (2) Is the colour of the 

 larva influenced by the colour of the environment which, acting 

 upon some sensory surface, directs the kinds and amounts of 

 pigments deposited, or absorbed from the food ? (3) It w.as also 

 important to test the effects of certain new food-plants and ot 

 others about which the evidence was conflicting ; and (4) to 

 look out for any indications which would throw light upon the 

 red-spotted varie'ies, or upon the existence of individual varia- 

 tion of any kind, under similar conditions of parentage and 

 food ; and (5) to inquire into the periods during which the larva; 

 are most susceptible to the colour-influence. 



Experiments in 18S5. — These were divided into five series, as 

 the larvre came from five batches of eggs. The differences 

 between the larva2 are expressed in five degrees : white, whitish- 

 intermediate, intermediate, yellowish intermediate, and yellow. 

 Series I. Eggs w ere laid by a moth bred from a lai-va which had 

 been a typical whitish variety (as was also the case with the 

 male parent). The resulting larvtewere whitish (5) upon /'yr-us 

 Mains (var. aecrlia), interjnediate (1) upon Populus Iremulu 

 and another species of poplar, intermediate (i) upon £alix 

 baliyloniea, and whitish (2), although they did not become adult, 

 upon 5. rubra and other similar species of sallow. In this 

 series the hereditary influence on the side of white is seen to be 

 strongly marked on comparing the effects with those shown in 

 the other series and in the parent larvae (see former paper). 

 Series II. Eggs were laid by a moth bred from a whitish-inter- 

 mediate larva, without any act of coitus having been vvitne sed 

 (although male moths were present). Most of the eggs shrivelled 

 up, but a few hatched, and form the larvaa of this series. Sub- 

 sequently coitus was induced (artificially), and a large number 

 of fertile eggs were laid which are considered under the next 

 series. The larva; of Series II. were whitish (4) upon Salix 

 viminalis, although not to such an extent as upon apple, and 

 yellowish-intermediate (l) upon S. Smithiana and another sallow 

 with similar leaves. These results are rather irregular, for the 

 former larvae were whiter than the parents, the latter yellower 

 than it is probable that the parents would have been on the saine 

 tree. But the results were not unusual in themselves. Series 

 III. The female parent was the same as in Series II. ; the male 

 parent was bred from a typical whitish larva. The resulting 

 larvEE were whitish (6) upon ordinary apple, and upon the same 

 whh the leaves sewn to expose the under sides (i), and to 

 expose the upper sides (3) (none of these reached maturity, 

 and the last lot were especially young when they died) ; whitish 

 (4) but immature upon crab (var. acerba) ; whitish-intermediate 



