306 
Transactions of the Royal Society of South Africa. 
described colour. My preliminary observations indicate that the band varies 
between \ 580 and \ 607 according to the reaction of the solution. 
Perhaps it is surprising that even " iodine-green " (heptamethylpara- 
rosaniline, with three methyls on one nitrogen) agrees with theory, as far as 
• can be seen. The value of X should be 593 x (0-972)-^ x (1-0245)7 or 645. 
The band observed in a commercial specimen was A 640. 
Another important point is that the effect of N-ethyl is scarcely dis- 
tinguishable from that of N-methyl — the value of (6) the y-lN -diethyl- 
amino-f actor being about 1'023 as against 1"021 for dimethylamino. 
A factor for the replacement of — OH by — JSfEto can be calculated 
from the foregoing data. The process of replacing y-OH by y-H has 
the factor 1 -^ 0-9657, and the factor for replacing y-H by y-NEtg is 
1-023 ; hence the required factor is 1*023 0*9657 or 1*061. Applying 
this to the case of fluorescein and rhodamine, we here use the factor with a 
different constant, viz. that of fluorescein, X494. The replacement of two 
hydroxyl by two diethylamino-groups should give a substance with wave- 
length 494 X (1*061)^ or 555. Observation on commercial rhodamine 
gave A 556. 
Correlation of the Phthaleins with the Aniline Dyes. — We now introduce 
a seventh factor, namely that of the orthocarboxyl group present in the 
phthaleins. This factor is very close to unity, being equal to 1*002. As 
examples benzaurine of ^ 553 may be compared with phenolphthalein 
X 554, and f '-tetrabromobenzaurine of A 583 with tetrabromophenol- 
phthalein of A584i. It will be sufficient at this stage to calculate only 
one or two of the observed colours. Phenoldimethylanilinephthalein 
(see Part II, p. 112) is the oxy-dimethylamino-carboxylic acid of fuchsene, 
and its calculated wave-length is 593 x 0*9657 x 1*021 x 1*002 or 587, and 
that was also the reading observed in the alkaline solution of this substance. 
Curiously enough, bis-dimethylanilinephthalein appears to be colourless. 
Its wave-length should be 622, as it is the carboxylic acid of malachite- 
green. The sulphophthalein, however, is not colourless, but has A 634 
or A 593 according as it is acid or neutral. 
At this stage the other factors for substitution in phenolphthalein may 
be combined with the parahydroxy-factor so as to express the compounds 
investigated in Parts I to YI (see the summary on the last page of Part 
lY) as direct derivatives of fuchsene. The ratio of phenolphthalein to 
fuchsene is 0*934. Now since the value of (8) the ortho- (or e-) bromine 
atom-factor is 1*0133 in phenolphthalein, the lengthy calculation of e-tetra- 
bromophenolphthalein from first principles as 593 x (0*9657)2 x 1*002 x 
(1*0133)4 may be shortened to 593 x 0*934 x (1*0133)4, both coming to A584, 
and agreeing with observation. 
We may again repeat here the remarkable circumstance that methyl^ 
ethyl, isopropyl, chlorine, bromine, and iodine in the e-positions have all 
