216 
Transactions of the Boyal Society of South Africa. 
free aminoazobeiizeue is not quiiione-imineplienylhydrazone, for in that case 
it would have X 332 (328 + 223 - 219) instead of \ 424. 
I have to confess that I have ahnost abandoned the hope of finding an 
explanation of the higher azo-dyes. For example, the following substances 
of widely different constitution have almost exactly the same absorption- 
band, and I hope that some other worker will attempt an explanation, now 
that the wave-leno-ths are available : 
- NH3CI (acid water) \ 497. 
>NH3C1 (acid water) \ 499, 
\0Na \ 370 
in water + 
NaOH, 
— ONa \ 492 in water + 
/ NaHCOa, 
iNaO— <^ ^— Nj— <^ ^—<( \— ONa \ 470 in water + NaOH^ 
whilst the other symmetrical substance 
NaO<^ ^— No— <^ N— No— <^ \0Na 
has A 515. 
Substitution of a naphthalene- for a benzene-ring has a larger effect in 
the azo series than in the di- or tri-phenylcarbinol series. Thus alkaline 
benzeneazo-a-naphthol has its band about 85 units above that of benzene- 
azoplienol, and the paranitro-derivatives (see Watson, I. c, p. 89, v ; and 91^ 
xiii) have the bands at \\ 494 and 576 respectively. 
Here, in conclusion, I may mention a most remarkable fact which some 
other worker maybe able to explain, if it is not a mere coincidence, viz. that 
the bands of (a) phenolphthalein, (b) phenol-a-naphtholphthalein, and (c) 
(di) a-naphtholphthalein are almost exactly in the ratio 10 : 11 : 12, viz. 
X\ 554, 608 and 662,* and those of (d) nitrobenzeneazophenol and (e) 
nitrobenzeneazo-a-naphthol are exactly in the ratio 6 : 7, viz. 494 and 576. 
This certainly appears to be connected with units of tautomeric change also. 
* The ratio agrees even better if 18 be subtracted from each for the C6H4C05H 
grovip, giving benzhydrol derivatives. 
