August 2, 19 17] 



NATURE 



445 



can do this by assuming the star to be in a steady- 

 state." This I cannot believe to be possible. 



Prof. Eddington now says that the star " must settle 

 into a state of density and temperature which would 

 produce an outward flow at the required rate." Per- 

 haps; but surely Prof. Eddington's original conten- 

 tion was that the rate of outward flow could not be 

 affected by density and temperature, but depended only 

 on the mass. J. H. Jeans. 



London, July 13. 



I AM in general agreement with Mr. Jeans 's re- 

 marks on the difficulty of obtaining a source of stellar 

 energy more powerful than the Helmholtz contraction. 

 It may be added that there is a conceivable source, 

 which was, 1 believe, once suggested by Mr. Jeans 

 himself, viz. a gradual annihilation ot matter by 

 positive and negative electrons occasionally neutralis- 

 ing one another. This would provide an almost in- 

 exhaustible store of energv', but there is the grave 

 objection that it affords no reason why the dense 

 dwarf stars should liberate so much less energy than 

 rarefied stars of the same mass. One would have 

 expected compression to be favourable to the process 

 of cancelling of electrons. The search for an additional 

 store of energy is not at all encouraging; but, on 

 the other hand, there are important arguments against 

 the short time-scale — notably Prof. Strutt's evidence 

 of the age of terrestrial rocks, and the time needed 

 for the tidal evolution of the earth-moon system. 

 I have not felt myself able to combat the arguments 

 on one side any more than on the other ; accord- 

 ingly, in the paper criticised by Mr. Jeans, the ques- 

 tion was left entirely open. In the one place where 

 it was necessary to consider the source of stellar 

 energ}", I attempted to show that my formula fairly 

 represented both the radio-activity and the contraction 

 hypotheses — having regard to the necessarily approxi- 

 mate character of the investigation. 



The opinions in the last paragraph of Mr. Jeans 's 

 letter seem much too sweeping. It is desirable to 

 criticise them, because his disbelief in my results is 

 presumably a corollary to his rejection of the possi- 

 bility of obtaining information from consideration of 

 a stationar\' or quasi-stationary state. If energy were 

 generated at a fixed rate within the star, the radiation 

 would no doubt have to take place at the same rate ; 

 but to bring this about the star must settle into a 

 state of density and temperature which would produce 

 an outward flow of energy at the required rate. We 

 have thus a triangular equation — generation of 

 energy = theoretical emission (def>ending on the trans- 

 parency and temperature-distribution) = observed emis- 

 sion (given by the effective temperature). Mr. Jeans 

 assumes that the imposed rate of generation must 

 necessarily be involved in any results that are derived. 

 But we can dispense altogether with the first mem- 

 ber, and obtain "new knowledge" from the equation 

 which remains. 



July II. 



I AM afraid we can scarcely trespass on your space 

 to enter on the detailed discussion which seems neces- 

 sary in order to arrive at an understanding. In a 

 paper about to appear in the June Monthly Notices 

 I have rearranged mr analysis in what is, I believe, 

 a more lucid form. If Mr. Jeans finds the result still 

 unsatisfactory, I hope he will renew the attack in 

 another place. 



With regard to his final point, I may give a word 

 of explanation It is true that I find that the total 

 radiation of a giant star depends only on the mass — to 

 my order of approximation. If a different rate of genera- 



XO. 2492, VOL. 99] 



tion of energy, fixed and independent of the density, 

 were imposed, the star could not settle permanently 

 in the giant state. If the supply were too sir^ll the 

 star would contract, though more slowly than on the 

 Helmholtz theory, and ultimately attain equilibrium 

 in the dwarf state. The case of too large supply 

 scarcely needs to be considered, since it involves an 

 evolution in the reverse direction from that geaerallv 

 accepted. This may, perhaps, be regarded as addi- 

 tional evidence of the difficulty of obtaining a long 

 time-scale by assuming an unrecognised source of 

 energy. A. S. Eddingtox. 



Cambridge, July 17. 



FORESTS AND RAINFALL. 



THERE are several questions regarding the 

 mutual relations of natural phenomena that 

 appear at first sight so simple that the obvious 

 answers may be received for generations as too 

 clear to require reconsideration. One of these is 

 the influence of forests on rainfall. It seems so 

 natural that if a large area of bare ground is 

 planted with trees which grow into a forest the 

 moisture of the district will be increased by in- 

 creasing rainfall, diminishing run-off, and, in hot 

 countries, falling temperature, that one scarcely 

 stops to inquire on what evidence the belief is 

 based. Everyone must remember the vivid picture 

 drawn in Marsh's "Man and Nature " of the deso- 

 lation wrought in Palestine and other Mediter- 

 ranean lands by desiccation consequent on the 

 destruction of forests and abandonment of cultiva- 

 tion. But in that work, as in most of the writ- 

 ings on this and cognate questions, the motto of 

 the discussion might he post hoc, ergo propter hoc. 



The problem has been attacked by innumerable 

 writers in Europe and America, and we do not 

 profess to have the mass of heterogeneous litera- 

 ture at our finger-ends. We do, however, retain 

 a general impression of unsatisfactoriness in the 

 methods and results, and the impression is re- 

 newed by the latest contribution to the subject, 

 the Indian Forest Bulletin, No. 33. This consists 

 of a "Note on an Inquiry by the Government of 

 India into the Relation between Forests and Atmo- 

 spheric and Soil Moisture in India," prepared by 

 Mr. M. Hill, Chief Commissioner of Forests of 

 the Central Provinces. Mr. Hill has presented 

 an admirable precis of what must be a large mass 

 of official documents, and he apf>ends t\vo excel- 

 lent memoranda by Dr. Gilbert Walker, the 

 Director-General of Observatories in India. That 

 the good work of Mr. Hill should leave an unsatis- 

 factory' impression is not his fault, but his mis- 

 fortune in having to deal with official reports in- 

 stead of plain scientific data. The history of the 

 investigation as set out in the bulletin is briefly 

 this : — 



In 1906 Lord Morley, then Secretary of State 

 for India, sent to the Viceroy a note from Dr. 

 J. Nisbet, formerly of the Indian Forest Service, 

 pointing out that "the relation of forests towards 

 the mitigation of the severity of famines " had 

 never been adequately considered. Sir William 

 Schlich forwarded with Dr. Nisbet's letter his 



