726 



EESEARCHES AND EXPERIMENTS. 



retains all its sensitivity for about two years, 

 provided it be sheltered from light and moist- 

 ure ; it is not affected either by heat or cold. 

 After the light-impression has been made on 

 the collodioiiized paper, the latter is first dipped 

 in common water, care being taken to make 

 the immersion complete. Here it must remain 

 for at least five minutes, or until the paper, 

 which was beginning to curl, becomes perfectly 

 flat. In the mean time the following solution 

 is prepared, in quantity only sufficient for the 

 pictures to be developed at once, for oftentimes 

 it decomposes in the course of a day or two : 

 distilled water, 1 litre ; glacial acetic acid, 20 

 grammes ; citric acid, 20 grammes ; pyrogallic 

 acid, 3 grammes. Into a basin with flat bot- 

 tom, and of a size corresponding to that of the 

 proof treated, is poured enough of this solution 

 to completely submerge the proof; a depth of 

 three or four millimetres is amply sufficient. 

 Into this is dipped the proof after taking it 

 from the water and draining it, the collodion- 

 ized side uppermost. After inclining the basin 

 in every direction so as to cause the liquid to 

 pass several times over the proof, a portion of 

 it is poured into a glass, and there a few drops 

 of the following solution are added to it : dis- 

 tilled water, 100 grammes ; crystallized silver 

 nitrate, 5 grammes. This is stirred well, so 

 as to mix thoroughly. The whole is poured 

 into the basin, which is again inclined as before. 

 The image now appears ; seven or eight min- 

 utes suffice to completely develop it, with the 

 sky or the lighted parts of an intense black. 

 When the proof is sufficiently developed, it is 

 put into water and then dipped in a solution 

 of hyposulphate of soda, 40 per cent., to fix it ; 

 it is then freely washed in water in the usual 

 way, and dried bet ween leaves of blotting-paper. 

 To obtain positive proofs, we detach from the 

 paper the layer of collodion, thus getting the 

 image on a thin transparent pellicle. The opera- 

 tion is very simple, consisting merely in adding 

 to the collodion layer firm and transparent sub- 

 stances until the cliche has attained the proper 

 thickness. To this end a normal collodion of 

 the following composition is prepared : gun- 

 cotton, 25 grammes ; sulphuric ether, i litre ; 

 alcohol of 40, litre. Lay the proof on a plate 

 of glass, having first turned up the edges all 

 round, so that the liquids to be poured upon 

 it shall not overflow. On the collodion layer 

 containing the image pour the normal collo- 

 dion, beginning at one ef the corners of the 

 proof most remote from the operator. Then 

 incline it slightly, so as to cause the liquid to 

 flow ; and, after the entire surface has been 

 covered, the excess of liquid is poured back 

 into the bottle. Then the clich6 is laid flat in 

 a roomy box, or in any other place where it 

 will be sheltered from dust, and left for a few 

 moments to dry. When fully dried, or when 

 ; is no longer sticky, we again pour over the 

 layer of normal collodion caoutchouc dissolved 

 in benzine. When this is dry, we apply a sec- 

 ond layer of normal collodion, then caoutchouc 



again, and lastly a final layer of normal collo- 

 dion. The cliche is now left to dry for twenty- 

 four hours. Then the paper is separated at one 

 corner by a finger-nail from the coating formed 

 on it, when it may easily be stripped off alto- 

 gether, leaving the paper clean and white. The 

 negative cliche" is at least as clear as though it 

 were of glass, but it has the advantage of not 

 being brittle, of occupying little space, of not 

 being spoiled by rubbing, and finally of giving 

 better proofs than can be got from glass cliche" s. 

 Composite Portraits. An ingenious method 

 has been devised by Mr. Francis Galton for 

 obtaining "composite portraits" that is, por- 

 traits (photographic) not of individuals sepa- 

 rately, but of aggregates of individuals, so that, 

 each one contributing an equal share, the result 

 will represent the sum of their features. In 

 order to make a composite portrait, in the first 

 place photographs of the faces which are to 

 make it up are collected. Suppose there are 

 eight of these photographs : they are carefully 

 laid one over the other, with the eyes as nearly 

 as possible superimposed. (Of course the por- 

 traits should all be similar in attitude and size.) 

 The portraits being thus arranged, a photo- 

 graphic camera is directed upon them. If, now, 

 it would require an exposure of say eighty sec- 

 onds to give an exact photographic copy of any 

 one of them, this time is divided among the 

 eight, each one being exposed successively for 

 the space of ten seconds. The sensitive plate 

 in the camera is then developed, and the print 

 taken from it is a "generalized picture" or 

 composite portrait. Those of its outlines will 

 be found sharpest and darkest which are com- 

 mon to the largest number of the components ; 

 the purely individual peculiarities leave little 

 or no visible trace. These individual pecu- 

 liarities being necessarily dispersed equally 

 on both sides of the average, the outline of 

 the composite is the average of all the compo- 

 nents. It is a band and not a fine line, because 

 the outlines of the components are seldom ex- 

 actly superimposed. The band will be darkest 

 in the middle whenever the component por- 

 traits have the same general type of features, 

 and its breadth or amount of blur will measure 

 the tendency of the components to deviate 

 from the common type. Mr. Galton points out 

 sundry ways in which these composite portraits 

 might be of use. One use, which will readily 

 occur to the intelligent reader, is that of fur- 

 nishing typical pictures of races of men. Thus, 

 if the photographs of a large number of indi- 

 viduals, taken at random from among the rep- 

 resentatives of a race, be in this way made to 

 form a generalized portrait, there is no doubt 

 that the distinctive race features will appear 

 with the minimum of individual traits. Another 

 use of this process is to obtain by photography 

 a really good likeness of an individual. The 

 inferiority of photographs to the best works 

 of artists, so far as general resemblance is con- 

 cerned, lies in their catching no more than a 

 single expression. But now, if many photo- 



