524 
PROFESSOR A. H. CHURCH OX TURACIX, 
The mean percentage deduced from the live series is 6‘92. If Series C be omitted 
the percentage becomes 6*82 ; if Series E, it is 7‘05 ; if both these series be excluded 
the percentage stands at 6*96. These three numbers are so near together that it 
seems hardly material which we accept, but after having given due weight to the 
various circumstances likely to affect the validity of the several results obtained, 
I incline to adopt the figure 6‘96. I may here mention that this percentage of 
nitrogen corresponds to 7'7 per cent, of copper, if the ratio N 4 : Cu be assumed, or to 
7 per cent, of copper in the case of the ratio N 9 : Cu 3 , 
§ G. Copper in Turacin. 
For the copper-determinations recorded in my first paper ( loc . cit., p. 632) I was not 
able, through scarcity of material, to employ adequate amounts of turacin. But the 
differences between my earlier results and those which will be given presently arose 
mainly from a strange and unsuspected peculiarity of turacin. When the pure dry 
pigment is quickly and strongly heated it gives off, as I mentioned in my previous 
memoir, “a violet vapour resembling that of iodine.” Now this vapour contains an 
organic copper compound, capable of being condensed into a red crystalline solid, 
which is soluble in ether and insoluble in alkalies—unlike the original body which 
yields it. If the turacin be slowly heated it gives off no visible or coloured vapours, 
even when the temperature is subsequently raised to a dull red heat. Here, then, is 
to be found the cause of the discrepancy between the six per cent, of copper found in 
my earlier determinations and the seven per cent, recorded below, In all the older 
analyses, the turacin, after having been dried at 100° C,, was heated alone in a 
crucible, whereby it lost about one-sixth or one-seventh of its metallic constituent in 
the form of the violet (or rather, purple) vapour mentioned above. When it was 
found that turacin, if first submitted for ten minutes to the temperature of boiling 
mercury, gave off no visible vapours when afterwards more strongly heated or even 
incinerated, an easy method for estimating both the copper and the accidental ash- 
constituents of this pigment seemed available ; accordingly, two quantitative experi¬ 
ments were carried out in the following way. The turacin was dried at 100° C. until 
constant in weight; it was then heated for one houi- in a bath of boiling mercury, and 
finally incinerated over a Bunsen burner until the weight of the residue became con¬ 
stant. The total ash was weighed, and then the copper present in it was determined, 
first, by titration with potassium iodide and sodium thiosulphate, according to 
Brown’s method ; and, secondly, by a gravimetric process, the iodine and the organic 
matter from the starch-indicator having been removed. The copper was precipitated 
as sulphide, and afterwards converted into and weighed as oxide. The total ash in 
the sample having been first of all ascertained in these analyses, an opportunity was 
afforded for the estimation of the fixed matter other than copper oxide left on the 
incineration of turacin. Here are the results :— 
