to the formation of other dyes by the decomposition of methylene 

 blue. Unna (1891) had already described what he called poly- 

 chrome methylene blue, made by heating a solution of methylene 

 blue on a water bath with potassium carbonate. Nocht decided to 

 use this in the Romanovsky stain instead of untreated methylene 

 blue. He found that it gave very good results if properly neu- 

 tralized before mixing with eosin; and then learned that better 

 results could be obtained by the use of a smaller amount of alkali 

 and a longer period of polychroming, without subsequent neutral- 

 ization. 



The next step in preparing blood stains was made by Jenner 

 (1899) who collected the precipitate formed when methylene blue 

 and eosin are mixed, and redissolved it in methyl alcohol. He did 

 not use polychrome methylene blue, and his stain lacked the 

 nuclear staining principle of Romanovsky 's and Nocht's stains; 

 but it was an important step in that he showed the possibility of 

 collecting the precipitated compound stain and of dissolving it in 

 some solvent other than water. Combining this procedure with 

 the Nocht stain was the next logical step and was taken inde- 

 pendently by Renter (1901) and by Leishman (1901). The method 

 thus introduced was briefly to follow Nocht's technic of combining 

 eosin with polychrome methylene blue, but then to filter off the 

 precipitate and to redissolve it in methyl alcohol, not adding fur- 

 ther water until the moment of applying the stain to the blood 

 films. 



Modern blood stains are in general modifications of Leishman 's, 

 differing only in detail. Wright's modifications, the one most used 

 in America, (see Mallory and AYright, 1924, p. 170) differs from 

 Leishman's only in that he prepared polychrome methylene blue 

 by heating for only an hour in flowing steam, whereas the Leish- 

 man technic calls for twelve hours at 65°, with subsequent stand- 

 ing for ten days. Balch's modification calls for a polychrome 

 methylene blue prepared by standing ten days with precipitated 

 silver oxide. 



Giemsa's and MacNeal's modifications are somewhat different. 

 Giemsa obtained methylene azure, in what he considered a pure 

 form, and combined it with eosin in order to obtain a more definite 

 compound than when polychrome methylene blue is used. Then 

 to obtain better differentiation he mixed it with methylene blue. 

 Following his instructions, the Griibler Co. put on the market a 

 product known as Azure II, which was a mixture of Azure I (i.e., 

 methylene azure) and methylene blue in equal parts; and also a 

 compound known as Azure Il-eosin, which was an eosinate of 

 Azure II, or more precisely a mixture of the eosinate of methylene 

 blue with that of Azure I, in equal parts. This latter compound is 

 the one generally known as the Giemsa stain. MacNeal (1922) 

 proposed a method for obtaining a very similar blood stain, pre- 



89 



