382 Messrs Purvis and Foster, The absorption spectra 
The absorption spectra of pyridine have been observed and 
described by Hartley ( Journ . Ghem. Soc., vol. 47, pp. 685 — 757); 
and Baker and Baly, more recently, in their valuable researches, 
have examined the spectra again, and compared the curve obtained 
with that of lutidine, a dimethyl derivative of pyridine, and 
also with the curves of various chlorine derivatives of pyridine 
{Journ. Ghem. Soc., vols. 91 and 92, pp. 1122 — 1132). 
The latter investigators showed that the effect of introducing 
two methyl groups into pyridine did not alter the character of the 
band, but moved it a little towards the red end of the spectrum, 
and also decreased its persistence. And, that, in the tri-, tetra-, 
and penta-chlorine derivatives of pyridine, the introduction of the 
chlorine atoms produced a gradual shift of the bands towards the 
red end, whilst the persistence was gradually increased. 
We have studied the absorption spectra of collidine and 
9-chlor-collidine in a similar way; and, in order to compare the 
curves obtained with those of pyridine and its derivatives in 
Baker and Baly’s experiments, we used equivalent strengths of 
N 
■ . _ - solutions in absolute alcohol, and plotted the results in the 
1000 
adjoining curve, so that the ordinates represent the logarithms 
of relative thicknesses, and the abscissae represent oscillation 
frequencies. 
As a result of this comparative study of the curves we find that 
the character of the collidine band does not appear to be different 
from that of pyridine or of lutidine, except that it is moved a 
little more towards the red end of the spectrum, whilst the 
persistence is decreased a little more than that of lutidine, and 
also, of course, more than that of pyridine. This result might 
have been expected because of the introduction of a third methyl 
group in the pyridine nucleus. The heads of the bands, ex- 
pressed in oscillation frequencies, are the following : 
Pyridine 3950 (Hartley) 
Lutidine 3750 (Baker and Baly) 
Collidine 3700 (Purvis and Foster). 
With regard to the 9-chlor-collidine, the character of the 
band is again not essentially different from that of collidine ; 
but there is a much greater shift towards the red end, and also 
a slight decrease in the persistence of the band. 
Now, the persistence of the bands of the tri-, tetra-, and 
penta-chlorine derivatives of pyridine described by Baker and 
Baly, is so much greatly increased by the introduction of the 
chlorine atoms into the nucleus, that it appears to be very 
probable the nucleus has not been greatly affected, if at all, by 
the introduction of nine chlorine atoms in the collidine molecule. 
