P/i(tt()ffrfij)/ii(' ItvHvnrche* ntul Portrait tire 329 



(F) Mt'fisnrements of ReHemhlaiice. 



T Iiave already referred to Gallon's long-continued researches on the 

 infiisurenit'iit of re'senihlance. lie f^ivo in Salyrr, Octolier 4, 1906', some 

 iicoouiit of his method and of his apparatus' for meitsurin^ his "index of 

 mistakahility." He opens his account with the following worfls: 



"At tlio distiiiu'c (if a fow scom'h of piicps tlio human faoc iipjx-ars to b«> a iinifonii rwldiHh 

 Miir, with nil Ncpiinito feiitures. On a nearer approtich spi-cks Ix-j^in to Ix" Hwn corrt'«|X)ndin>{ to 

 (lie eyes and mouth. These ){ra(hially incre.t«e in dixtinctnefis, until - -at about thirty |tfice.s — iho 

 features heoome so clear that a hitherto unknown person could thereafter tx' recognised with 

 some aN.Hurance. There in no Ijetter opportunity of observing the eflectii of diNtancc in con- 

 foundin)^ human faces than by watchinjf soldiers at a review. Their drew i.s alike, their pose 

 is the same, the light falls up<in them from the same direction, and they are often immoveable 

 for a considerable time. It is then noticeable how some faces are indistinguishable at distances 

 where great diversity is appaii'nt in others, and the rudelydefine*! idea will be justified that 

 the (listAiK-e at which two faces are just mistakable for one another might serve as a trust- 

 worthy basis for the ineasiireinent of re.semblimce. The same may be said of obscurity, of con- 

 fu.sed refi-actions, and of turbid media." 



In the apparatus descrihed in this paper in Nature, Galton used distance. 

 But he also looked at two portraits through a graded series of "blurrers*," 

 or i^la.sses with different thicknesses of Canada Balsam placed upon them. 

 Finally he adojjted a neutral coloured wedge (like the wedfje photometer used 

 tor star magnitudes), lookinj^ at the portraits through thicker and thicker 

 parts of it until they were "mistakable." The apparatus is fairly simple for 

 the distance olxservations. There is a six-foot ba.se ooard upon which are two 

 sledjfes carried along its length by endless cords each g<»ini; round their own 

 |)air of wheels, one at either end of the board. At the summit of the ba.se 

 l)oard, which slopes slightly downwards from the observer, is a screen with 

 an eye-.slit to carry sjioctacle lenses for examining the photographs; it can 

 l)e replaced by a bracket ujjon which optical combinations can be mounted 

 for throwinf^ the photographs to a consideiable distance, i.e. greater than 

 that of the ba.se board, in the manner of an inverted telescope. The sledges 

 each carry a stjvndard to which the portraits to be examined can be attached, 

 and when attached they can be rotated in azimuth to conipen.sate for differences 

 of degree in the photographs of "half" face. The position of tlie photographs 

 with regard to the observer's eye can be read on a scale which runs down 

 the centre of the ba.se lK)ard. 



Galton's procedure is as follows: He first mejisures in millimetres the 

 distances a and u' from the pupil line to the lip line of each porti-ait. He 

 then takes from the ba.se board .scale the distances </ and d of the two 

 ])ortraits from tlie eye screen in centimetres. If now the indices n = lOOu/il 

 and n'=lOOu'/d' be formed, then when they are equal, the two portraits 

 subtend the satne angle at the eye, and this allows for any difference in size. 



' Vol. i.x.\iv, pp. .562-3. 



' Now with a good many additions, devised by Galton him.self, in the Galton Laboratory. 



' He presented me in 1907 with a series of "blurrers" and there are a good many sets in 

 the Galtouiana, but I have not come across any ivccount of their preparation and standartl- 

 isation. The photometric wedge is a much more permanent measurer. 



Poii 48 



