March 26, 1908J 



NA TURE 



485 



sample of seed which had been sent from Senaar Province. 

 This sa iiple was sown at Ghizeh, and the material sent 

 to Kew was taken from two adjacent plants in the same 

 row, these being plants developed from embryos which 

 had ripened on a tree in Senaar. The reason for sending 

 material from two different plants was because this row- 

 showed two distinct leaf-forms, some of the plants having 

 much wider leaf-lobes than the others, and these two types 

 were represented by 213-1 and 213—2. In no other respect 

 could any distinction be drawn between the two types, at 

 least on ^y character within my grasp ; moreover, they 

 all flow-ered within the same week, ripened within the 

 same week, grew to a remarkably uniform height, and 

 had similar habits of growth ; with the one exception of 

 the leaf-shape they were far more alike than a similar 

 group of plants taken from a field of any variety of 

 Egyptian cotton. These two forms were separated by 

 Watt into G. arboreiun and G. Nanking, because " a 

 y-ellow-flowered G. arboreiun with deeply laciniate 

 bractpolcs and three glands on the under surface of the 

 leaf would destroy the specific isolations " (p. 138). I 

 incline to think that the iaciniation of the bracteoles and 

 the glandulation of the leaves should have been made the 

 subject of comparative study — in order to ascertain their 

 capacity for fluctuation — before such erratic characters 

 were entrusted with the responsibility for this violent 

 separation of the tw'o forms into two separate species. 

 Such comparative study would at least have been com- 

 menced had these plants been seen growing side by side 

 in my plot. 



On the other hand, we find on p. iSi that Moqui Indian 

 cotton from Arizona (209-3) ^^^ " Hindi " weed cotton 

 of Egypt (55 A) are placed together under G. piinctatum. 

 Waiving the query as to why Hindi, a naked-seeded cotton, 

 should be placed in the fuzzy-seeded section, I should like 

 on other grounds — but in all diffidence — to advance the 

 opinion that if the two strains could be grown together 

 at Kew, or examined side by side on my plot in Egypt, 

 they would be systematically removed from one anotlier 

 by a wide interval. 



The employment of common names has also been men- 

 tioned by Colonel Prain ; the following instance, therefore, 

 does not seem altogether pointless : — the plant referred to 

 as 56.C.2 (p. 224) came from a sample of Afifi cotton, and 

 bears lint of the brown Afifi colour ; this colour is the 

 characteristic and sole morphological distinction of Afifi 

 from Abbassi, the latter bearing white lint, so that 56. C. 2 

 could by no possibilitv be described legitimately as " close 

 to Abbassi or Afifi." 



The cultivated varieties of Egyptian cottons — and prob- 

 ably of Uplands — consist of many different strains mingled 

 together and cross-fertilised, resembling one another in a 

 few obvious ch .racters of economic importance. Thus, on 

 p. 224, Watt describes the strain 142, plant A, as being 

 distinct from the Abbassi plant described in par. 2 of the 

 same page. In point of fact, No. 142 was taken from a 

 prize sample of Abbassi. 



Though I wish to see an exact method adopted for the 

 investigation of this labyrinthine genus, such method to 

 be based on pedigree culture and statistical inquiry, I 

 am nevertheless grateful to Sir George Watt for having 

 gathered together the mass of detailed information which 

 is to be found in his book, and I hope — with Colonel 

 Prain — that we shall not have to wait long for the publica- 

 tion of further researches on the subject. 



W. Lawrence Balls. 



Cairo, February 27. 



In the courteous letter in which Mr. Balls exercises his 

 right to criticise details in Sir G. Watt's work on cotton, 

 as to which he considers himself a competent judge, he 

 gives expression to some misapprehension that it may be 

 wr>ll to remove. 



It has not been affirmed that the ideas of the writer 

 of the review which appeared in Nature for January 16 

 as to " species " and " varieties " do not accord w'ith 

 accepted usage. What it was necessary to point out was 

 that the reviewer had not made it clear that his inter- 

 pretation of these words accords with accepted usage. 

 There are two passages in the review in which the words 

 arc dealt with together ; in one passage they are so used 



NO. 2004. VOL. 77] 



as to imply that the status of a variety is the same as that 

 of a species ; in the other they are so used as to indicate 

 that a species is subordinate in status to a variety. The 

 ideas of the reviewer may be as precise as those of Mr. 

 Balls ; they may, on the other hand, be as loose as his 

 own phraseology ; he has given us no means of deciding. 



The position assumed by me has already been explicitly 

 stated. I have reserved perfect freedom of judgment as 

 regards the acceptance of Watt's conclusions, not as to 

 the limits of species in the genus Gossypium alone, but as 

 to all the issues involved. When he explains that his 

 general position is the reverse of this, it will be felt that 

 Mr. Balls does himself an injustice. ■ 



The name of the distinguished public servant referred 

 to by Mr. Balls is Mr. A. F. Broun, and is not as given 

 in Mr. Balls' letter. D. Prain. 



The Isothermal Layer of the Atmosphere. 



In his letter in Nature of February 27 Mr. Dines asks 

 why the adiabatic conditions wliich prevail in the lower 

 part of our atmosphere should suddenly cease at a height 

 of about 40,000 feet. The answer comes more readily if 

 the question is altered to, Why does the isothermal con- 

 dition of the outer layers of our atmosphere suddenly cease 

 at about 40,000 feet ? The isothermal condition or even 

 increased temperature with height is the condition which 

 would naturally prevail in an atmosphere surrounding a 

 smooth sphere. For if the sphere is a very hot one its 

 entire gaseous envelope should acquire its temperature, 

 whereas if the solid sphere, like our earth, is cold, and if 

 heat from the sun is warming the atmosphere by radiation, 

 one may expect the outer layers to be warm and the lower 

 layers to be the coldest ones. If, however, there are 

 irregularities, as, for instance, mountain chains on the 

 earth's surface, then the air, whenever it is forced over 

 them, parts with its moisture as it rises on the one side 

 and then descends on the other side as a dry and hot 

 Fochn, in which wind the conditions are perfectly adia- 

 batic, the temperature gradient rising steadily with de- 

 creasing height. It seems, therefore, that it is our 

 mountain ranges which prevent the isothermal condition 

 from descending below the height at which effective mixing 

 or moisture removing occurs. 



This leads to the conclusion that if at one time ou.' 

 mountain ranges were lower than at present, the iso- 

 thermal condition and its low temperature will also have 

 been lower than at present. This may have been the case 

 during Glacial periods. On the other hand, during tropical 

 periods our mountain ranges may have been higher than 

 they are at present ; the isothermal condition will have 

 ended at a higher level, and the steady rise of temperature 

 below this boundary will have resulted in a \-ery high 

 temperature on the earth's surface. 



I remember discussing this subject about twenty years 

 ago at Ai.\-la-Chapelle with Dr. A. Ritter, who had only 

 recently in Wiedemann's Annalen (vols, v.-viii., " Heights 

 of Atniospheres and Conditions of Nebulae ") dealt with it 

 very exhaustively. If I am not mistaken, it was the Foehn 

 wind which had first led to these inquiries, but, strange to 

 say, Dr. Ritter relied on molecular motions for the neces- 

 sary mixing of the layers. This may have been due to his 

 feeling that if isothermal conditions were conceded, an 

 interstellar atmosphere would have to be postulated. We 

 therefore almost naturally disagreed as to the possibility 

 of condensing the so-called permanent gases, which fact 

 had not then been accomplished. My view was that if 

 nitrogen and oxygen should be condensable, and if the 

 adiabatic condition existed up to the outer limits of our 

 atmosphere, then, at the zero temperature to be found 

 there, both gases would condense and sink to the lower 

 levels, to be followed by further and further layers until 

 the whole atmosphere would be deposited on the earth's 

 surface. Dr. Ritter merely pushed this difficulty further 

 awav bv saying that, even if oxygen and nitrogen could 

 be condensed, our atmosphere might nevertheless be 

 surrounded bv hydrogen. Now that hydrogen has been 

 condensed, helium would have to take its place, or, and 

 this is a view not easily accepted, our earth may be 

 surrounded by a very attenuated and possibly warm inter- 

 stellar atmosphere. I think that the recent experiments 



