366 SCIENCE PROGRESS 



which three carbon prints are made of the required colours from 

 the three colour record negatives and then superposed. 



As any method by which the three coloured impressions can 

 be superimposed on paper will give a three colour print, various 

 printing-press methods are available, and some exceedingly fine 

 work has been done by collotype printing, the printing surface 

 being of hardened gelatine and inked up with printers' inks. 

 But the process that is of all the most advantageous, especially 

 where many impressions are wanted, consists in the preparation 

 of three " half-tone " blocks, one corresponding to each of the 

 three colour records, and making from these three superposed 

 impressions in the three necessary colours. The greatest 

 difficulty of these methods is in the getting of suitable inks, 

 for they should not only be of the correct colour, but they 

 (or at least two of them) should be transparent, and they should 

 be permanent. The most suitable colouring matters are too 

 fugitive, so theory has to be compromised, and as none are really 

 transparent there is always a tendency for the colour last printed 

 to predominate. In spite of these drawbacks very excellent 

 work is done commercially by this process, but probably never 

 without "fine etching" — that is, proof impressions from the 

 blocks are compared critically with the original by persons who 

 have been trained to appreciate colour and drawing ; and where 

 the tints are incorrect the alterations required are marked on the 

 impressions, and the blocks are returned to the etcher, who 

 works on the faulty part. In the best work this proofing and 

 correction is repeated until the result is satisfactory. The " half- 

 tone " block prints a series of dots equidistant from their centres 

 and larger as the tone or shadow is deeper. In superposing the 

 three impressions no attempt is made — indeed it would be prac- 

 tically impossible — to print the corresponding dots exactly over 

 each other, and the result is that sometimes the dots are super- 

 posed and sometimes juxtaposed, according to chance. This 

 might appear to lead to a mixture of the additive and subtractive 

 effects, and as the colours for the one method must be the 

 complementaries of those used in the other, it might be supposed 

 that confusion would result. It is not so, however. Where 

 only one ink is required, and where all three are required to 

 their fullest extent to give black, the case is simple. Take the 

 intermediate case of a grey produced by juxtaposition only of 

 the three inks and in such a manner that the white paper 



