Oct. 30, 1884J 



NA TURE 



(>2>7 



year. It leads to a reflection on the inevitable incomplete- 

 ness of a catalogue. There is no pause in the publication 

 of books. In spite of the most careful filling up of the 

 lists of missing books by the librarian, and the most 

 liberal expenditure by the Committee, hundreds of new 

 books must have come out, and a large proportion of them 

 added to a library, between the time when the last title is 

 handed to the printer, and the time when the first out- 

 sider can purchase his catalogue and examine what are 

 the treasures kept in store for him. And in no production 

 of industry, not even in ladies' adornments, is novelty so 

 important a recommendation as in literature. The dis- 

 heartening reduction of prices in secondhand catalogues, 

 not of three-volume novels only, but of laborious and im- 

 portant works, is a proof of this. A greedily read daily 

 press makes it inevitable. Any printed catalogue, there- 

 fore, with all the books in due order, must be deficient of 

 the favourite, if not of the most important books which 

 the library contains. Catalogues therefore in general 

 should be printed li ;e the most fugitive of literature, and 

 be renewed as frequently as possible. A card-catalogue 

 alone can be kept on a level with the stock of books. A 

 frequent publication by a large library of a list of its new 

 purchases, sold at a remunerative price to students and 

 luxurious readers, would make a library popular among 

 those with a strong appetite for reading, while it would 

 not lead to the older tenants of the shelves being forsaken 

 by the crowd. 



In most public libraries an effort is made to combine 

 the functions of the old collection of books with that of 

 the dispensary of useful or pleasing thought, by having 

 two departments. The books more deserving of the old 

 feeling of preservation are wisely placed apart with real 

 works of reference to form the Reference Department. 

 A mischievous result of this arrangement usually is 

 that it makes books of greatest intrinsic value and for- 

 bidding costliness least available to the impecunious 

 student. The Halifax Catalogue avoids this by arranging 

 all together in one alphabetical list, marking each of the 

 reference books with an /?, and leaving the question of 

 lending them out practically to the discretion of the 

 librarian or Committee. We strongly approve of this 

 method and of liberality in working it, and recommend it 

 to the notice of other libraries. W. ODELL 



THE "WENT/SCOPE" 



T T appears from the Pall Mall Gazette of October 21 

 *■ that there is a prospect of "a campaign being run 

 in the country " on behalf of the " Claimant" by " six of 

 the best orators whom money can collect, . . . supplied 

 with a hundred identiscopes." These are optical instru- 

 ments, containing on the one side a drawing made from 

 a portrait of the undoubted Roger Tichborne, and on the 

 other side a drawing made from an equally undoubted 

 portrait of the Claimant taken nineteen years later, and 

 the arrangement is such that on looking into the instru- 

 ment the drawing; combine into one. This, it is main- 

 tained, leaves no doubt that the two portraits are those of 

 one and the same individual. 



The more important of the questions raised by this 

 announcement is whether the fact of two genuine portraits 

 blending harmoniously into a single resultant is stringent 

 evidence that the portraits refer to the same person. Those 

 who have examined the optical combinations and photo- 

 graphic composites that 1 have exhibited at various times 

 will know that this is not the case. Those who have not 

 seen them and care to know more about the subject should 

 look at my " Inquiries into Hum in Faculty." (Let me take 

 this opportunity of correcting an error there. The full and 

 profile composite libelled " two sisters," in the middle of 

 the upper row of the frontispiece, is really one of three 

 sisters. I had made many composites of the family, and 



by mistake sent the wrong one to the printer.) The reason 

 why photographic portraits blend so well together is 

 that they contain no sharp lines, but only shades. The 

 contour of the face is always blurred, for well-known 

 reasons dependent on the breadth of the object-glass ; 

 even the contour of the iris in an ordinary photographic 

 print looks very coarse and irregular when it is examined 

 by a low-power microscope. On superimposing a second 

 portrait, the new shades fall in much the same places as 

 the former ones ; wherever they overlap they intensify 

 one another ; where they do not overlap they leave a 

 faint penumbra which has usually a soft and not un- 

 pleasing effect. Judging from abundant experience, there 

 would be no difficulty in selecting photographs of many dif- 

 ferent persons that should harmonise with the photograph 

 of the Claimant, and it would be amusing to try strange 

 combinations. I could suggest one that I think would 

 succeed excellently : it is of a certain distinguished member 

 of Her Majesty's but I must be discreet, though pro- 

 bably if I ever come into possession of suitable photographs 

 I may make a private experiment. 



It seems, however, that the identiscope is not intended 

 to be used to combine reproductions of the actual photo- 

 graphs, but only drawings in bold lines that have been 

 made from them. The photographs, it is to be presumed, 

 do not agree in aspect, so drawings are made from them 

 that do so, the diameter of the iris being used as the scale 

 unit of the breadth and length of the features, in making 

 the drawings. Although the diameter of the iris is 

 spoken of as an invaluable unit for exact reduction, its 

 disadvantages appear to be great : (1) Its vertical dia- 

 meter was, I suppose, not used, because in the large 

 majority of cases the upper part of the iris is covered 

 by the eyelid. (2) The horizontal diameter is un- 

 available unless the eye of the sitter was directed 

 straight at the camera ; otherwise the iris is seen in 

 perspective, and its breadth is reduced by an unknown 

 amount. (3) One eye is perspectively larger than the 

 other, unless the face was set truly square to the optical 

 axis of the lens ; if not, it would be necessary to measure 

 both eyes and to take a mean ; this is a requirement to 

 which I have as yet seen no allusion. (4) The diameter 

 of the iris is only about 1 '25th part of the length between 

 the chin and the vertex of the head, consequently any 

 minute error in its measurement would be largely multi- 

 plied when applying it as a unit. (5) The diameter of the 

 iris in a photographic print does not, as I have already 

 implied, admit of accurate measurement. The identi- 

 scope appears to be the same as an instrument sold some 

 years ago, and of which I have one now by me. The 

 description printed on it is " E. Wolf and Sons' patent 

 Limnoscope, for copying drawings, designs, &c." I bought 

 it for the purpose of experiments with composites, and 

 tried many modifications of its principle, but other plans 

 proved so much better that I discarded it. The principle 

 is easily realised by any one who cares to place a table 

 by a closed window and then to go out-of-doors with an 

 open book in his hand, which he must hold horizontally 

 by the side of the window, at the level of the table. He 

 will then see through the glass an image of the book (a 

 " Pepper's ghost," in short) resting on the table. The 

 reflected image is so faint that the direct image has to be 

 dimmed. Yellow glass serves this purpose. The limno- 

 scope is not suitable for combining ordinary photographs 

 because the reflected portrait is reversed ; the left side 

 of one face is combined with the right of the other. 

 Much better instruments exist for making optical com- 

 binations ; I have described them in my book. 



I conclude as follows. First, that the fact of two photo- 

 graphic portraits blending harmoniously is no assurance 

 of the identity of the persons portrayed. Secondly, 

 when drawings made from portraits are shown to blend it 

 does not follow that the portraits from which they were 

 drawn would blend equally well. And lastly, the photo- 



