284 



KNOWLEDGE, 



[December 1, 1896. 



in France large quantities of an excellent burning and 

 cooking oil are made from the nuts. 



Little or no vegetation will flourish under beech trees ; 

 the carpet of dried and decaying leaves and husks will bo 

 found much the same all the year round, seldom enlivened 

 by anything grcon. This is the result of the dense shade, 

 and the light, dry nature of the soil in which the tree 

 delights, being drained yet more ihoroughly by the closely 

 matted roots spread near the surface. Two remarkable 

 edible fungi, however, grow well under beech trees : these 

 are the morel and the trutlle. Both arc much sought after 

 in France and Clermany, being highly prized for the table. 



Several interesting varieties of the beech are in culti- 

 vation. The best known are the purple beech — originally 

 discovered in a wood in Germany — the copper-coloured 

 beech, and the fern-leaved beech, which has its leaves cut 

 almost like a fern. 



HELIUM AND PARHELIUM. 



By E. Walter Maunder, F.R.A.S. 



N the April Number of Knowledge an account was 

 given of the very beautiful researches which led 



I Profs. Eunge and Paschen to the belief that 

 the gas which Prof. Ramsay had extracted from 

 cleveite was not a simple substance but a mixture of 

 two distinct elements. The force of their conclusion may 

 be better appreciated if we glance at the adjoining 

 diagrams, wherein the spectrum of cluveite gas is analyzed 

 in the same manner as the spectra of lithium and sodium 

 were in the earlier paper." In the first diagram the 



* The formulfe employetl by Profs. Runge and Paschen, Astro- 

 physical Journal, January, 1896, for the six series of the cleTeite gas 

 are as follows : — 



Helium : — 



Principal .. ^= 38455:i2-4. 



SuborJinate I. ^= 2922435 



II. \= 2919796-7 

 Parheliuni : — 



Principal .. ^= 3202986 



Subordinate I. ^ = 2717516 



11. ' = 2716859-5 



ro98919x 

 1-098363 X 

 1-061524 X 



1 095 37 X ' 

 1-097587 X 



— 1-4507 X 



— 1-67 X 



— y'656 X - 



-I- 1 9636 X 



— 2-726 X 



1-088256 X-,. — 3-596 x 



It "ill be noticed that the third power of » is used instead of the 

 fourth, as iu the formula pixen on page 88. Probably tlie exact 

 formula would contain an indefinite number of terms, with increasing 

 powers of n. 



The lines in the two spectra are as follows, the wave lengths 

 being given in tenth-metres, the wave numbers in oscillation 

 frequencies to the metre ; — 



complete spectrum of the cleveite gas appears in the first 

 line, except that double lines are shown throughout as 

 single, the scale of the diagram not permitting the members 

 of the doublets to bo shown separately. The second line 

 shows the complete spectrum of the parhelium constituent ; 

 the third, fourth, and fifth lines, its resolution into its three 

 series ; the sixth and three following linos show the complete 

 spectrum of the hehum constituent, and its similar reso- 

 lution. 



There is a point to which it was not possible to refer in 

 the earlier paper and which should be noticed. Of the six 

 rhythmical series of lines into which the cU'veite gas was 

 completely resolved, two approached one limit, two 

 another, and were hence regarded as subordinate series. 

 Of the other two — from their greater brightness and wider 

 stride evidently principal series — one was taken as being 

 associated with one pair of subordinate series, the other 

 with the other. But in default of any known relation 

 between principal and subordinate series, this association 

 might have been as reasonably reversed. There is, how- 

 ever, an intimate relation which clearly marks out how in 

 this instance the spectra are connected. 



Referring to the April paper, we find that the limiting 

 wave number of the principal series of lithium is 

 4351930, and the mean of its two subordinate series 

 2862672. The difference of these gives 1489258, very 

 close to the wave number of the first line of the principal 

 series. So with sodium : we have for the limit of the 

 principal series 4149634, for the mean limit of the 

 subordinate series 2452128, and for the first hne of 



Helium. 



N. 



Wave Length. IWaveNnmber 



Pakhelium. 



Wave Length WaveNmiiber. 



The lines marked with an asterisk are double, each having a faint 

 companion on the redward side, with wave number on the average 

 100-7 smaller than that of its primary. The line mai-ked with a 

 dagger is possibly double, the companion liaving a wave length O'OS 

 larger than that of its primary. 



