May 3, 1894J 



NA rURE 



li. as against 1 158 in vol. i. Under Arcliidce the author 

 Deludes the following groups as sub-families, which have 

 I usually been treated as families by previous authors : — 

 \Arctiinie, LithosiiiKe, iKyctcolincE, and Nol/ncv. The Aj^ar- 

 ' isiti/ie a.Te a small family of handsome day-flying moths, 

 and certainly look rather out of place in the position 

 which they occupy in this book. The extensive family of 

 NoctuidiC is divided into ten sub-families i'lrijimc, 

 Aconliina, PaUndiince, Sarrothripiitcc, Eitteliina:, Stictop- 

 Jerina, Gonopterina-, (2!/adriJinte, FocilliniT, and Deltoi- 

 dince), of which the two last are held over to the forth- 

 coming third volume of the book. 



Concernincj the A'oi/ia'dir, Mr. Hampson remarks, 

 " The lowest forms are those of which the larvae have five 

 pairs of abdominal prolegs, and the perfect insects have 

 vein 5 of the hind wing fully developed, and from the 

 centre of the discoceliulars, this ancestral form being only 

 found in some Deltoidiitcc and SarrothripincB." 



As the plan of the second volume is identical with that 

 of the first, which we had the pleasure of noticing in 

 Nature for February 23, 1S93 (pp. 387-388), we need 

 only add that there seems no lalling oft" in its e.\ecution. 

 It is hoped that the third volume, including the 

 Epicopiidcr, Uraniida:, Epiptemidit-^ and Geometn'dtB 

 will be completed in the course of the present year. 



W. F. K. 



LETTERS TO THE EDITOR. 



[7'he Editor Joes not hold himself responsible for opinions ex- 

 pressed by his correspondent:. Neither can he undertake 

 to return^ or to correspond :cith the writers of^ rejected 

 manuscripts intended for this or any other part o/NatuRF. 

 A'o notice is taken of anonymous communications.'] 



Panmixia. 



Mr. Rom A.N ks has requested those sUidents of natural 

 history who cannot accept the doctrine of Panmixia to show 

 the error which they believe to lie in his reasoning. I therefore 

 ask leave to explain why I am unable to accept either the first 

 proposition put forward by Mr. Romanes in Nature of to-day, 

 or the doctrine itself. Mr. Romanes says : — 



The sitrz'izal-mcan must {on cessation of selection') fall to the 

 birth-mean, A;c. This statement involves neglect of a way in 

 which selection may, and often must, operate. i\ simple 

 example will show this. The mean height of adult Englishmen 

 is roughly 67J. inches ; and if I offer to enroll in a regiment 

 every adult Englishman who i.s more than 66 and less than 69 

 inches high, the mean height of my regiment will, as every 

 Statistician knows, be still 67^ inches, but I shall be obliged 

 to reject more than half the population. .\ form of selection, 

 involving the destruction of more than half the population, may 

 thcicToic occur without affecting the mean value of the character 

 I hope shortly to publish evidence, based on the 

 .'cTiX. of many thousands of animals of one specie?, at 

 !• iity s'oges of growth, showing that selection does in fact 

 I'l trate in this way in particular cases. That it must so operate 

 n. many cases is obvious from the fact that many wild animals 

 n main for several generations without sensible change in their 

 njcan character. In these cases either selection acts as I 

 suggest, or it is incapable of affecting a change in the mean, or 

 it does not act at all. 



The second and third propositions put forward by Mr. Romanes 

 are not demonstrated by any statistics with which I am ac- 

 quainted ; and with regard to the extreme statement that " any 

 lailure in the perfection ol hereditary transmission will be weeded 

 out " by selection in a wild slate, I would urge the need, which 

 has lately been well pointed out by Bateson, of a quantitative 

 measure of the efficiency of selection. The frequency of even 

 considerable abnormalities in specialised organs of wild adult 

 animals, of which so many admirable examples are described 

 in Mr. Batcsons recent work on variation, show, if it needed 

 showing, that natural seltclion is in most cases an imperfect 

 agent in the adjustment of organisms. 



But my main difficulty is that neither Mr. Romanes, nor Prof. 

 Weismann, nor any other advocate of the doctrine, has shown 



NO. 1279, VOL. 50] 



that in some given case Panmixia does in fact occur, and that 

 the results predicted are in fact produced. On ihe other hand, 

 Mr. Gallon has shown that civilised Englishmen are themselves 

 in a condition of Panmixia, at least with respect to several 

 characters, especially stature and the colour of the eyes. Now 

 the mean stature of Englishmen is known to be slowly increas- 

 ing, and there is no evidence of the disappearance of coloured 

 eyes. 



My objections to the position of -Mr. Romanes and others are 

 therefore two : tirst, that it is based on the assumption that 

 selection, when acting on a species, must of necessity change 

 the mean character of the species — an assumption incompatible 

 with the maintenance of a species in a constant condition ; and 

 secondly, that in the only case which has been experimentally 

 investigated, the consequeoces said to result from a condition of 

 Panmixia do nor, in fact, occur. \V. F. R. Weldon. 



University College, London, April 26. 



On Some Sources of Error in the Study of Drift. 



Asa general rule we may feel sure that the boulders scattered 

 over the .^urlace of a district which consists chiefly of boulder 

 clay, have been derived from the underlying deposit. There 

 are, however, some cases in which the inference is unsafe. For 

 instance, the Thames now marks the southern limit of the 

 glacial drift — a curious circumstance, and one of which a wholly 

 satisfactory explanation has not been given. Many think that 

 this sharp definition of the southern limit of the glacial drift 

 is so improbable that they would fain attribute some deposits 

 in Nonti Kent to the glacial period, or at any rate would 

 expect to find a few sporadic boulders stranded on the slopes 

 of the North Downs ; and there far-transported fragments do 

 not unfrequently occur. 



But there is this great source of error. All along the lower 

 Thames barges carry refuse and rubbish of every description 

 from London, and this is taken, such as it is, and laid on the 

 adjoining lands. 



So you find carried on, with road scrapings, fragments of 

 every kind of road metal ; with soil turned out in digging 

 foundations, specimens of all the materials used for building ; 

 with the contents of middens, every variety of object of 

 domestic use or ornament. It is marvellous what large lumps 

 get on to the land in this way. When, then, anyone produces a 

 specimen, even a large specimen eight or ten inches in diameter, 

 and perhaps taken out of a deep loam, the evidence is rejected. 

 The stone may have been carried on to the land with the 

 manure, and the loam may in that district be quite recent rain- 

 wash. It may be that some of them were really ol glacial 

 origin, out all are equally distrusted. Some of them certainly 

 cannot be referred to ice action. I have seen large pieces of 

 Napoleonite found on the sur.'ace in North Kent. By what 

 accidents they came to be there we cannot tell, but we may, at 

 any rate, acquit the ice of having had anything to do with the 

 transport of that peculiar Corsican rock. 



When walking along the base of a cliff of boulder clay, we 

 may generally infer that the far-travelled boulders that lie at 

 its base have just been washed out of it. In most cases they 

 have been ; but in some, and those often the cases in which it 

 is of greatest consequence to have the origin of the boulders 

 clearly established, we have another serious source of error, of 

 which I have just seen a good example. 



A Norwegian vessel, carrying timber from Chrisliansund to 

 Boston, in Lincolnshire, ran aground and became a total wreck 

 off Old Hunstanton last winter. 1 saw her in January. The 

 vessel looked sound enough to a landsman's eye ; but she was 

 dismasted and gutted, and the salvage was on the sand dunes 

 close by. About her a pool of varying brcidth had been 

 foimed by the swirl of the water round the hull. The currents 

 had been deflected by various circumstances here and there, 

 as especially where a quantity of ball.ast had been thrown out. 

 This consisted of large boulders of various kinds of gneiss and 

 porphyry, and the weighty pile looked as if it were little afliected 

 by the currents of the incoming and receding tides. 



In April, I visited the spot again, expecting to find that the 

 boulders had been driven along the shore by the fierce storms 

 which had raged along that coast since my previous visit, and 

 inteniliiig to make note of their dispersal and the distance to 

 which they had travelled. I found, however, that the keel and 

 a portion of the lower part of the wreck remained, and that 

 the surrounding pool was greatly deepened and extended. 



