May II, 191 1] 



NATURE 



3t>: 



Satisfactory definition is obtained throughout the whole 

 length (43 cm.) of the plate used in the spectrograph 

 camera. In the more refrangible portions of the spectrum 

 Lumiere " Sigma " plates were employed, while from 

 \ 4900 to the red end Seed " Gilt Edge 27 " plates sensi- 

 tised by Wallace's formula were used. For comparison 

 the spectrum of the iron arc was photographed alongside 

 the star spectrum. 



Six plates of the spectrum of Sirius (A 4200 to X 6600), 

 four of the spectrum of Procyon (A 4200 to \ 4900), and 

 nine plates of the spectrum of Arcturus (A. 4300 to A 6600) 

 were available for measurement, and the chief object in 

 the study of these spectra has been to test the possibility 

 of detecting any differences of displacement for the 

 different lines, and thereby obtain some idea of the effective 

 pressure in the atmospheres of the stars. In Sirius the 

 number of lines available for measurement was compara- 

 tively small; in the case of Procyon and Arcturus the 

 selection of lines was similar to those used in the investi- 

 gation of similar displacements of the lines in the spec- 

 trum of the sun's limb. 



The enhanced metallic lines, it will be remembered, show 

 as a class most definitely larger shifts at the sun's limb 

 than the ordinary arc lines. Now in the spectrum of 

 Sirius the enhanced lines form a prominent feature, while 

 the arc lines are few. In Procyon the enhanced lines are 

 less prominent, while the arc lines have become more 

 pronounced. In Arcturus the enhanced lines are almost 

 evanescent, while the arc lines, which are associated with 

 the spectrum of sun-spots, are verj' strongly developed. Mr. 

 Adams gives a table showing in summary form the main 

 results of the inquiry, from which it is seen that in all 

 cases the enhanced lines show a decided displacement to 

 the red relative to the arc lines. Giving the displacements 

 as radial velocities m kilometres, we may summarise the 

 results as : — 



Sirius: Enhanced — Arc Lines= +090 km. = +o*oi4 A 

 Procyon: ,, ,, ,, =+0'5S ,, =+0*009 



Arcturus: ,, ,, ,, = +0'o8 ,, = +0"OOI 



The behaviour of the prominent lines in Arcturus is so 

 definite that a special discussion is given of them. A 

 large proportion of the lines of titanium, vanadium, and 

 calcium are greatly strengthened, the enhanced lines 

 decidedly weakened, and those of iron and chromium 

 either strengthened or weakened according to their tempera- 

 ture gradation. The lines of nickel appear to be more 

 prominent in the star spectrum than in sun-spots. The 

 following table summarises this discussion : — 



Klement displacement \XdIy"' 



H 



Ca 

 Mg 



V 

 Ti 



Ni 



Fe 



A 



-0-O2O 



-o'oi7 



-O'OII 



-o"oo6 

 -0006 

 -o-oo6 



+ 0006 



km. 

 - 12 



-070 



-0-68 

 -0*24 



-023 



-0"22 

 + 0-25 



Th'' -^liifts evidently suggest definite grouping of similar 

 clrin-ni^. The iron lines show a shift towards the red 

 c(ini]).ii ■'I with all the other elements examined. 



Surh is the material Mr. Adams provides for his in- 

 vestigations. In the absence of any other known probable 

 cause, he considers pressure as the principal agent causing 

 these systematic displacements in stellar spectra. The 

 laboratory experimental results of Humphreys and others 

 gave as an average shift for the arc lines of iron 00025 A 

 per atmosphere of pressure. At the sun's limb the 

 enhanced lines in the more refrangible portion of the spec- 

 trum were found to be shifted approximately 50 per cent, 

 more than the arc lines, and recent work by Mr. Gale on 

 the spectrum of titanium indicates that the enhanced lines 

 of this substance arc also shifted more than the arc lines 

 at the same pressure. Assuming, thon, that a similar 

 relationship exists between the enhanced and arc lines of 

 other elements, this affords a means of estimating the 

 gravitational pressures in the atmospheres of stars the 

 spectra of which show these displacements. Thus, as seen 

 in the table quoted above, the enhanced lines in the spcc- 



NO. 2167, VOL. 86] 



trum of Sirius are shifted towards the red relative to the 

 arc lines by 0-014 A. This would correspond to a pressure 

 of 12 atmospheres in excess of that existing in the sun's 

 reversing layer. Similar reasoning in the case of Procyon 

 indicates a pressure of 7 atmospheres over that of the sun's 

 reversing layer. These results appear to be in accord with 

 the modern view of regarding stars of the Sirian type as 

 possessing no true photosphere, being simply a mass of 

 gas increasing in density towards the centre without any 

 surface of discontinuity or condensation. In such a star 

 the light coming from great depths would most probably 

 be visible from outside, and indications of great pressure 

 would then be expected. In Procyon the spectrum 

 indicates a transition stage between Sirius and the sun, 

 and the pressure is shown intermediate also. It should 

 be noted here, however, that one of the most important 

 cases investigated by Humphreys in his work on pressure 

 effects is directly opposed to the above conclusions. He 

 found that in the case of calcium the blue g line was 

 shifted by pressure about twice as much as the H and K 

 violet lines. Now the behaviour of these lines in the 

 laboratory, and also in the spectrum of the solar chromo- 

 sphere, indicates that H and K are typical enhanced lines, 

 while 4226 ig) is a very typical arc line. The differential 

 pressure effect on the enhanced and arc lines of strontium 

 was exactly similar to that of calcium, viz. the enhanced 

 lines were shifted less than the arc lines. Unfortunately 

 for this discussion, Humphreys only employed the arc 

 spectrum in his pressure investigation, so that the general 

 behaviour of the enhanced lines of other substances than 

 calcium and strontium cannot be inferred from his results. 

 Passing on to the conditions of pressure in Arcturus, it 

 is pointed out that the facts indicate the existence of a 

 well-formed photosphere, the light from which proceeds 

 from relatively low-pressure areas at moderate depths. 

 The results for the lines of different elements indicated in 

 the table are similar to those found for the solar lines. 

 Thus in the sun hydrogen rises to very great heights, 

 calcium and magnesium also being high-level substances 

 Titanium is also relatively high-level, but iron is distinctly 

 a low-level element. In Arcturus the displacements indicate 

 exactlv such an arrangement, and it is thus concluded that 

 the liries of H, Ca, Mg, Ti, &c., are subject to less pressure 

 than those of iron, and therefore that the gases producing 

 them lie at a higher average level. 



Charles P. Butler. 



RECENT INVESTIGATIONS ON SOIL 

 FERTILITY. 



Ih OR some years past the United States Department of 

 -*■ .-Xgriculture Bureau of Soils has maintained that 

 infertility might, and not unfrequently does, arise from the 

 presence in the soil of toxic organic substances that have 

 been excreted from the roots of plants. This view has been 

 opposed on two grounds : it is not evident that plants 

 do normally excrete poisonous substances ; and if such 

 substances are present there is no proof that they would 

 act as poisons in the soil, which possesses a remark- 

 able power of withdrawing dissolved substances from solu- 

 tion. Not long ago Scheiner isolated dihydroxystoaric 

 acid from a considerable number of unproductive soils, 

 and now, in conjunction with J. J. Skinner,* he has 

 examined its behaviour to plants in water cultures. In 

 all cases its effect was toxic, but the toxicity w.is much 

 reduced when fertilisers were added to the solution, and 

 was at a minimum when the fertilising constituents were 

 present in the ratio most favourable to plant growth. 

 Several incidental questions were also cleared up dealing 

 with water cultures— perhaps the most difficult of all ex- 

 periments to interpret— and the paper contains a great 

 number of data bearing on the subject. The behaviour of 

 this acid in the soil is not touched upon, and very wisely 

 no attempt is made to argue from a water culture to a 

 soil. It is. however, a distinct step in advance that an 

 acid has been isolated from certain soils and identified, 

 and shown to be poisonous in water culture. The results 

 may well be connected with the known fact that, in 

 absence of lime, soil becomes acid and loses fertility, which 

 can onlv be restored by addition of lime or chalk. 



1 Bulletin 70, Bureau of Soils, U.S. Dep.-xrtment of Agriculture. 



