104 



NA TURE 



[May 30, 1S95 



In habit and si/t .ui >^v.dling is not at all like latiata, but 

 might be taUen lor a poor specimen of the common Cinerarias. 

 In several characters it is intermediate l>etween latiata and the 

 latter. The stem is rather woody, less so than in latiata, but it 

 is thick like those of garden kinds : |)etioles like latiata in 

 having no auricles : leaves, nevertheless, large like those of 

 garden kinds, the backs vcr)' woolly, but largely purplish, as in 

 many cultivated sorts. Now this plant must be either (i ) a 

 si).jrt from latiata in the direction of the garden forms, or (2) an 

 accidental hybrid between latiaia and one of the cultivated kinds 

 growing in the same house with it (we have no otners). The 

 latter seems more likely— an opinion in which Mr. Lynch fully 

 concurs. 



Similarly Bouch^ ( Wiltin. Motiatss. xxii. p. 298, orig. not 

 seen, quoted from Focke, I'fl. Mischlitige, 1881, p. 201) says 

 that a hybrid between C. U'ehhii (Schlz. Bipont.)and iiiienta 

 arose in the Berlin Botanic Garden as the spontaneous product 

 of these S|)ecies grow ing side bv side. 



It was, I think, on evidence like this that the parentage of the 

 older hybrids was conjectured ; but that Drummond and Men- 

 ders m certainly— and possibly others — did make definite efforts 

 to hybridise, cannot on the evidence be doubted. That these 

 efforts went no further than the brushing of |x>llen of some 

 species upon the flowers of others, I fully believe, and that on 

 such endence xhe precise parentage cannot be assigned is obvious. 

 Nevertheless distinct seedlings resulted. In a few years, as 

 the writer in Pa.xt. Mag., 1842, p. 125, says (in an article urging 

 t3 fresh efforts in crossing), this hybridisation " was the means of 

 creating quite a novel and su|)erior race." There were the new 

 jilants : how had they arisen ? Those who doubt that these 

 new kinds were hybrids must choose the other horn of the 

 dilemma, anJ accept them as sports pure and simple. 



That the historical records may contain errors, I am fully 

 aware ; but if they cannot be accepted in detail, should they be 

 altogether rejected ? We might perhaps reserve a doubt whether 

 the King came precisely from pure iiiieiila fertilised by latiata : 

 whether ctiieina var. laitea was a hybrid between (riienta and 

 fopiilifolia (as ile Candolle surmises) ; whether ll'aterlimisiaiia 

 was the offspring of true criieiila and true tiissilagiiiis : whether 

 Mrs. I^udon's s:atcment that the species used were ct-uenta, 

 Janata, aurila, liissilagitiis, antl popiilifolia, or Moore's belief 

 that cnieiita and tiissilasitiis, w ith jwrhaps Hhitieri ( = lanata), 

 moiier'Tiisis { = aiirita), and popiilijolia ( " Cross and Self-Fert.,"' 

 V- 335' >">!')> Of Otto's similar declaration ( Kegel's Garten/iota, 

 1857, p. 66), or that of the /our. d'hort. Gaud, 1S46, already 

 given, should each be taken without hesitation :ls full and com- 

 j)lete statements of the whole truth, but that they contain a 

 substance of truth is hardly in (juestion. 



Against this Mr. Dyer offers nothing but an opinion derived 

 from an inspection of certain modern plants. He who has con. 

 fidence in the results of this methoil must suppose our knowledge 

 of the laws of inheritance to be nmch more complete than I 

 lielieve it to be. It is not the method Darwin used. Take a 

 well-ascertainc<l case. Who would know from ins|)ection of the 

 Himalaya rabliit that it came directly from the Silver-greys or 

 Chinchilla.s ? (See Animals and Pits., i. p. 113.) It is unlike 

 (hem, is of .sudden origin, and yet breeds true.' To sup|x>se 

 that in cross-bred offspring given characteristics of the parents 

 must be found, is to a.ssume the precise question which in a dis- 

 cusMon of organic stability is at issue. Lei it be remembered 

 that on the hy|»thei>is of hybrid origin for our Cinerarias it is 

 sup[K)sed that they result from several species and varieties, 

 crossed not once only, but many times, in wholly irregular ways. 

 Can it be seriously expected that any s|>ecial resemblance to a 

 given ancestor should Ix' still traceable.'- 



My |x.»ilion then is this. We heard Mr. Dyer's statement ; 

 turning to the literature I fouml an entirely different account, 

 l«irne out by cipious and on the whole fairl) attested evidence, 

 |><>inling irnsisiiMy to the conclusion that the Cinerarias are 

 .s|)ecies which liybridi-e freely, and that our modern forms have 

 arisen through such hybrid unions. 



' T.Mr r.in..., rr.,r. ,),,.:„ K, A. ,!..,..;.•• r.i-! i;r.iv.r. and to hi. fore^ 



■■'jictially for (Ik 

 ' - in hU Hlr.iiii 

 t un. Afl-r this 



Mr. Djer has well s.iiJ that "if you take any statement that 

 Mr. Darwin has put forward, you may feel assured that behind 

 it is a formidable body of carefully considered evidence not 

 likely to be upset." By the courtesy of an opponent I have been 

 directed to a ])ass.ige in "Cross and Self-Fertilization," 1876, 

 p. 335' "here (before describing experiments .showing consider- 

 able self-sterility in the garden Cineraria) Darwin gives this 

 definition of his material, '" Senecio (riientiis (greenliouse 

 varieties, coiiinionly called Cinerarias, probably derived from 

 sei'eral fruticose or herbaceous species much intercrossed"). It 

 seems, therefore, that in this matter also Mr. Darwin has, to use 

 Mr. Dyers words, "squeezed out'' of the evidence "all that it 

 would profitably yield." 



Here I would fain leave the subject. But perhaps it may be 

 suggested that though Darwin's Cinerarias were proliably hybrids, 

 our Cinerarias may not be their descendants. Such a suggestion 

 involves the supposition that in some hidden pl.ace there was a 

 thin red line of pure crucnta waiting for the moment wheii it 

 should oust the hybrids. If this be seriously suggested, I shall 

 ask where such a strain was kept, an<l what steps were taken lt» 

 ]>reservc its purity. In view of the evidence that chance bleml- 

 ings occur freely, to keep a pure strain w<tuld require some care. 

 Until this has been proved, we shall not, I think, be wrong in 

 sup|3osing that each grower worked on the material his 

 preileces.sors had created, and that our Cinerarias are the lineal 

 descendants of the hybrids raised in the first half of this 

 century. 



In the course of this discussion, Mr. Dyer has treated me 

 to some hard words, which I do not particularly resent. 

 Whether I have deserved them is not, perhaps, for me to judge. 

 But I will ask Mr. Dyer to point out when, on being asked for 

 the facts upon which I ha\'e basetl a view, I have replied thai 

 that was a " matter for future collection." The facts 1 have 

 been able to collect may be few, but by a study of the w ritings 

 of my antagonists, I have not been able to add materially to 

 their number.' W. Baieson. 



St. John's College, Cambridge, May 26. 



It has been pointed out to me that my remarks on Mr. 

 Bateson's account of the Cineraria have been interpreted in a 

 sense of which I did not dream when I wrote them. 



I wish, therefore, to say that, although I do not believe Mr. 

 Bateson's reading of the jiassages I quoted to be the true one, 

 yet I have never questioned his sincerity in suggesting it, and 

 1 am pained to find that I have seemed lo do so. 



.M.ay 24. W. 1'. K. Wki.don. 



:ir !.luirai.ui .-.liuultj .i.*»l;rl iU'jil' lu tin 



CAutUiun 



of 



I 'tMr. Dyer on hu own ground, I have a»umed, what 1 



'_.^rr._! .iJmii, ili.it in none of llic varliju* niodern strain* traces of the 

 dtfTcrent parcnt-^pecic* appear. 



NO. 1335. VOL. 52] 



Boltzmann's Minimum Function. 



I (;.\tllKK from Mr. Culverwell's last letter (Naiikk, ;\|)ril 

 18), and Mr. Bryan's (.May 9), that we may regard the follow- 

 ing conclusion as establisheil, namely, the proof of the H theorem, 

 for any system depends on a certain condition (.V) being fulfiUeil 

 among the coordinates and innmenta of the molecules forming 

 the system. Consiilering these as clastic spheres, and using Dr. 

 Watstm's notation,///, . . . dii^ is the chance that a sphen- 

 shall have for cmmlinates and momenta /, . . . Pi+ap^, &c., 

 and Fi/F| . . . i/Qj the chance that another sphere shall have 

 I'l . . . I'l + rfPi, kVc. The condition required is that /' ami K 

 are independent, even for two spheres on the point of collision. 



Otherwise we may express it. Let there be /; spheres in 

 sp.ace .S. Let us suppose .Mr. Culverwell lo .i.ssign to each its 

 position at time / = o, and Mr. Br)an to assign independently 

 to each its coHi|v)nent velocities. Then the comlilion A is 

 fulfilled when / = o. 



,/H . 



We can then prove that when / = o is negative, or, as 



Herr Boltzmann would have us say, is more likely lo beneg.itive 

 than positive. 



Now arises a question which .seems to me to deserve con 

 siileration. A.ssuming our system lo be finite, and to be left to 

 itself unaffected by external disturbances, does il necessarily 



1 It has l>ecn inip«>ssil)lc f.ir nic to incorporate in this letter all the nULss of 

 informalitni whi«:h has Ireen nimt Kencrously sent nie liy correspoiidents 

 since this controversy Iwtjan. It is sugKesled that I should point out that 

 Mr. I)yer"s use of the word "feral" t.» mean "wild" is iu»t usual. A cor- 

 respondent tells nie that it w:ls proliahly first used in the special sense of 

 " run wild " by ll.unillon Smith, ^'a^. Lihr., Mammalia, iSjg, ix. p. 9». 

 Il has since been soused hy nuiny authors, especially D.irwin, ./<i. ami 

 I'llt., i. p. 117. &c. 



