122 Proceedings of Royal Society of Edinburgh, [sess. 
it differs from colocynthin. The green matter is best prepared 
by agitating the dry pulp with ether, evaporating down and dis- 
solving the ethereal residue in strong spirit. In this manner the 
hitter principle and fatty matters are eliminated.” The solution 
closely resembles a solution of chlorophyll of equal strength. 
Both are beautifully dichroic, appearing green in thin, and red in 
thicker layers; both also have a red fluorescence. The chief difference 
on inspection is that a much thicker layer of chlorophyll is 
required to yield the red colour by transmitted light. 
Spectrum . — The spectroscope at once shows that this green 
colouring matter is not chlorophyll. Spectra (a), ( b ), and (c), in 
the plate are for different thicknesses of the trichosanthes colouring 
matter, while ( d ) gives the spectrum of chlorophyll from cabbage 
of a thickness and strength corresponding as nearly as possible 
with ( b ). Taking this as the most characteristic spectrum, it may 
be thus described: — The spectrum begins at 1736, and slight 
shading continues to 1693. The first band begins (penumbra) at 
1638 (W. L. 654*1), and ends about 1600 (W. L. 615), the 
maximum absorption being from 1631 to 1627, and from 1605 to 
1602. From this there is a small amount of absorption till the 
second band begins at 1576 (W. L. 593*4), and continues to 1543 
(W. L. 566*8), the maximum absorption being from 1570 to 1563. 
From this there is no perceptible absorption till the third band, 
which begins at 1516 (W. L. 548*4), and continues to 1495 
(W. L. 534*8), the maximum darkness being from 1511 to 150L 
There is a fourth band, very faint, with its centre about 1452 
(W. L. 510*6), and a fifth extending from 1400 (W. L. 485) to 
1371. The spectrum ends about 1355 (W. L. 467). The relative 
depths of the maxima of these five bands are 10, 5, 4, 1J, 1. 
When a thin layer of the solution is used the spectrum shows 
two strongly marked bands from 1625 to 1612 (darkness 10), and 
from 1569 to 1550 (darkness 4), and a slight trace of a third band 
with its centre at 1501. The spectrum ends sharply at 1340. 
With a very thick layer the first band widens out in both directions, 
and strong absorption — sufficient to mask bands II. and III. — con- 
tinues to 1485. There is also much increased absorption in the blue. 
How completely these spectra differ from that of chlorophyll will 
best be seen from the plate. The chief band of chlorophyll lies 
