February 13, 1896] 



NA TURE 



341 



So far as I know, there does not now remain a single mathe- 

 matician or astronomer who favours a purely astronomical theory 

 of an Ice .\ge ; a theory which, as Arago, Humboldt and Croll, all 

 urged long ago, is quite inadequate to explain the climatic effects 

 required. Every one, as far as I know, now agrees with the 

 American astronomer Meech, who subjected the astronomical 

 theory to a most searching analysis, as far back as 1857, that 

 " the causes of notable geological changes must be other than 

 the relative position of the sun and earth under their present 

 laws of motion." It is with this sentence that I close my own 

 analysis of the problem in chapter ix. of my " Glacial Nightmare." 

 As I understand. Sir R. Ball in surrendering his old view, 

 which was that astrononiual causes by themselves are sufficient 

 to produce an Ice Age, falls back upon a modification of 

 CroU's meteorological argument. While, however, he no longer 

 relies on the adequacy of astronomical causes alone as com- 

 petent to produce an Ice Age, he does not admit the conclusive- 

 ness of Mr. Culverwell's argument, but bids us remember that 

 the world cannot be cut up into a number of parallel zones 

 shut off from each other by solid partitions, each one of which 

 can be treated as a separate climatic region, but that the 

 climate of every zone is very largely indeed the result of heat 

 brought in or carried away by air and water from or to other 

 zones. No one disputes this. It is in fact an elementary 

 postulate of meteorology, and applies as much to Sir R. Ball's 

 arbitrary zone termed a hemisphere as to any other. 



What we want Sir R. Ball to do is not to rest content with 

 this barren postulate, but to apply it as Croll applied his 

 postulates, and to prove that, granting the greatest possible alter- 

 ation of the relative length of the seasons due to eccentricity, <S:c., 

 which, as Mr. Culvervvell has shown, will by itself tend to shift the 

 climate of each zone about five degrees, how is this going to 

 affect the circulation of the air and of ocean currents sufficiently 

 to constitute an Ice Age ? This was the problem Croll virtually 

 set himself to analyse^ by a minute and ingenious investigation. 



Croll's arguments have been riddled through and through by 

 several writers, and in this behalf I may perhaps venture to 

 again refer to a minute dissection of them in a chapter, headed 

 "Transcendental Meteorology," in the work already cited, 

 namely, the "Glacial Nightmare," and which I have been told 

 by some eminent physicists is unanswerable. I can, at all events, 

 say it has not been answered. 



If Dr. Ball can discover some method of curing the radical 

 defects of Croll's arguments, he will have made us a valuable 

 present. Meanwhile, if I do not entirely misunderstand his 

 present }X)sition, it is more clear than ever that he owes it 

 to us all to withdraw his " Cause of an Ice Age " from circula- 

 tion, for it has not only been condemned by its distinguished 

 and formerly friendly critic, but has been actually condemned by 

 its own author. Henry H. Howorth. 



Athemcum Club, January 30. 



The Positions of Retinal Images. 



I'rok. Kuli'E, in his " Outlines of Psychology" (translation 

 by Prof. Titchener), sets out with much effectiveness the argument 

 in favour of believing that the visual perception of extended 

 surface is an original datum of consciousness attached to the 

 extended retinal surface (and no more to be explained than why 

 the sensation red feels the way it does, and not otherwise) ; and 

 he also shows conclusively that the peculiarity of nerve-excitation 

 by which right- and left-ness and up- and down-ness are dis- 

 tinguished, is of peripheral (and not of central) origin ; by 

 adducing the facts of metamorphopsia, that is, the cases in which 

 a portion of the retina has become detached by a wound, and 

 has afterwards grown on again, and in which vision is corre- 

 spondingly inverted — exactly as when a piece of the skin of the 

 forehead has been grafted upon the nose, say, and upon touching 

 it we seem, for a long time afterwards, to be touching the fore- 

 head. He thus attaches himself to the innate-spacesensation 

 theory of James and Sumpf. But his effort to show that the 

 out- and in-sensation is fundamentally dependent upon the 

 different .shape of the image cast upon the two retinas by an 

 object, carries less conviction with it. This is, of course, an 

 essential element of the sensation when the object looked at is 

 so complex as to consist of two points at a given distance from 

 each other. But when it consists of a single bright point only, 

 we are still perfectly able to determine its position in depth (if 

 it is looked at with two eyes), and the sensation-element which 

 enables us to do this is plainly moxt. fundamental than the other. 



NO. 1372, VOL. 53] 



To say the least, it is something which ought not to be over- 

 looked. 



It is plain that in this case the only criterion which is left us 

 (granting, what is the case, that the localisation can be effected 

 with certainty with two eyes, but only vaguely and indefinitely 

 with one) is the distance apart of the double images ; it is that 

 which we estimate, unconsciously of course, in spite of the fact 

 that one image is in one eye and the other in the other, and it is 

 that which we translate, without difficulty, into a feeling of depth. 

 But there is always an ambiguity ; for everj' point, O (Fig. i), 

 without the horopter-circle, which casts images upon the retina 

 at the points r and /, there is a congruent point, O', within the 

 horopter-circle, which casts images upon the corresponding retinal 

 points, r and r , and which, therefore, gives images which are 

 at the same distance apart. We have no difficulty in sensation 

 in distinguishing between a bright point at O and one at O', but 

 how can this be effected ? There is still a difference in sensation 

 between the two cases. The nasal half of each retina gives 

 distinctly brighter images than the temporal half; in the case of 

 the object O, which gives the two images, r and /, the remoter 

 one is the brighter, while in the case of the object O', which 

 gives the images / and r', it is the nearer one which is the 

 brighter. A bright image of the object, which seems to us to 

 be the thing itself, is attended by a somewhat fainter secondarj 

 self, whose presence we are absolutely unconscious of, in our non- 

 scientific lives, as an image,' but which we evaluate with the 

 utmost nicety as a sign of the distance away of the real object, 



and which has, moreover, a different significance according as it 

 stands nearer to, or further from, the fixation point than the image 

 which we regard as the object itself. This explanation may 

 seem at first to vary much in the air, but its correctness has been 

 demonstrated by Schon in a very ingenious manner (Arch. f. 

 Ophthalm., xxii. and xxiv. ). His experiment has been un- 

 accountably overlooked by all the makers of text-books, as far 

 as I have seen, but it is of critical importance. He arranges a 

 series of screens with openings in them in such a way that two 

 different bright objects are seen, one by the right eye only, in 

 the line h.^O produced, and the other by the left eye only, the 

 line //j O produced. The positions of the double images now 

 correspond equally well for an object at O or at O' ; and the 

 person experimented upon thinks he sees an object now at O and 

 now at O', exactly in accordance with the way in which the 

 relative brightness of the objects beyond the screens is made to 

 vary. When the image which falls at r is brighter than that 

 which falls at /, the object is seen at O ; when the image which 

 falls at / is brighter than that which falls at r (sufficiently brighter, 

 of course, to counteract the relative efficiency cf the different 

 halves of the retina) the object is seen at O'. It is therefore 

 demonstrated that it is the relative brightness of the images 

 which is the determining factor in enabling us to localise objects 

 in one or other of the two congruent worlds without and within 

 the horopter-circle. I have myself repeated the experiment with 

 perfect success. C. Ladd Franklin. 



Baltimore, January 13. 



J There .-ire many people who cannot bring the econdary ima^e into 

 consciousness, no matter how hard they trj-, when t falls at any distance 

 from the fovea. 



