398 Mr Gustav Mann on the 



light falls on the slit of the spectroscope ; direct sunlight 

 should be used in the first instance, and if a thick layer of 

 threads has to be examined, the rays should be condensed 

 by a lens. By arranging the test-tube in such a way that 

 the light falls through its lower end, we get the following 

 spectrum (vide spectrum 1, Plate XL): — • 



1. In the middle, between the lines B and C, a rather 



dark band, \ 678 — X 662. 



2. To the right of C a second very faint band. A, 654 



— X 638. 



At the blue end of the spectrum absorption commences at 

 X 531, and then shows again very little absorption about 

 X 459, thus we have marked out : 



3. Absorption to the left of X 459, the centre of this 



absorption about X 485. 



4. To the right of X 459 absorption again occurs, and 



I think it is possible to see most absorption al^out 

 X 450, the spectrum to the right of X 450 (about 

 X 445) showing again less absorption, after which 

 we have the normal absorption towards the 

 extreme end of the spectrum. 



This result, so different from the absorption-spectrum 

 usually given, inasmuch as band 2 has, to my knowledge, 

 never been observed before, made me at first dubious about 

 the material I was working with, tlierefore some fresh 

 material was got from the pond in the Eoyal Botanic 

 Garden and investigated by the same method, and it gave 

 results identical with the first. It appears, therefore, that 

 what Kraus has described as the first band of chlorophyll 

 really consists of two bands ; and in confirmation of this, if a 

 thick layer of Spirogyra is used in the experiment by letting 

 the light fall through the upper part of the test-tube, prepared 

 as was stated above, only one band will appear in the red 

 end of the spectrum, and this band has a breadth X 686 — 

 X 641, including, therefore, the two bands, one with wave- 

 lengths X 678 — X 662, and the other with wave-lengths 

 X 654 — X 638. The second band (X 654 -X 638) "^can 

 only be seen, firstly, if the layer of green material examined 

 is thin enough to prevent band 1 becoming very evident by 



