ON COLOURING MATTER OF BONELLIA VIRIDIS. 167 
Its body is globular and lies in a large cavity in the rock, 
whilst along fan-like proboscis is exserted from a fissure con- 
nected with the cavity. Consequently Bonel/ia is not exposed 
to sunlight. The green colouring-matter occurs in clusters 
of fine granules dispersed in the tegumentary tissues, and is 
apparently deposited in the protoplasm of the epidermic 
cells. The green colour is of so deep a tint that the animal 
appears to be nearly black. The drawing of the spectrum 
of this substance when in the living animal, made by Mr. 
Lankester, agrees so closely with what can be seen in the 
alcoholic solution, which I have studied, as to show that no 
decided change has taken place, although it has been kept 
exposed to the air for more than five months, and thus it 
appears to be a compound of very considerable permanence. 
The position and relative intensity of the bands shown in the 
drawing also differ no more than was probably due to the 
colouring matter being associated with some oil, and not 
dissolved in the solvents used in my experiments. 
In all my previous publications I have expressed the posi- 
tion of the absorption-bands seen in the spectra of coloured 
substances by referring them to a scale furnished by the 
black interference bands seen in the spectrum of the light 
passing through a plate of quartz between two Nicol’s 
prisms, the thickness of the plate being such that the 
Fraunhofer line D is 34, and the line F 7, or thereabouts. 
I still use this scale for actual observation, but it appears to 
me that for the future it would be far better if all writers on 
such subjects would express the position of the absorption by 
giving the length of the waves of light at that particular 
part of the spectrum in millionths of a millimétre. In 
order to be able to do this I have constructed a table of the 
wave-lengths of every part of my quartz scale, so that, after 
having measured the position of any absorption-bands, I can 
immediately express it by numbers representing millionths 
of millimétres of wave-length. The advantages of this 
system are that, not only is it a scale that all may adopt for 
general comparison, but, as I shall be able to show in some 
subsequent paper, most important relations can be shown to 
exist between the wave-lengths of different bands, which 
relations cannot be recognised if any other scale be adopted. 
I propose, therefore, in future to express the position and 
character of bands by millionths of millimétres of wave, and 
by printing under the numbers symbols to indicate the rela- 
tive intensity of absorption, thus: 
