TU.VXSIMRATION" OF PL.VNTS. 87 



of the inost-rcfraugible half of the spectrum. As this glass 

 not only transuiits yellow audj green which coiitaiu the fainter 

 chlorophyllinn bands II., III., and IV., but the powerful one 

 No. 1. as well, one would, on a priori grounds, infer that jnore, 

 and not less, transpiration would occur with yellow glass than 

 with red ; but such is not usually the case. Moreover, the loss 

 of water under ordinary clear colourless glass (one twelfth of an 

 inch in thickness) is sometimes less than under cither red or 

 violet glasses alone ; so that from these facts one is led to infer 

 that the ])resence of the brightest or yellow rays is an actual 

 impediment to transpiration, or in some way hinders the acciou 

 of the red and violet. The thickness of the yellow^ glass used is 

 one twelfth of an inch. 



Crreen, This excludes red and violet rays, but transm'ts 

 liglit which includes the position of the cliloru2)hyllian bands III., 

 IV., Y., and VI. It is one twelfth of au inch in thickness. 



BJtce. This has the red end greatly subdued, appearing quite 

 black in diffused li^rht : but in direct sunlio^ht a broad black band 

 is seen in the midst of a small quantity of dull red. It includes 

 green; but a black band in the position of the chlorophyllian 

 band No. IV. Hence the 2:lass transmits liir^^t vv'hich would 



include the bands No. V., \i,, and VII. The thickness of the 

 glass is one tenth of an inch. 



Violet. This passes less green, but rather more red, than the 

 blue glass. The light transmitted would include bands Nos. II., 

 v., VI., and VII. A dark absorption-band occurs about the 

 position of No. IV. 



Clear, Ordinary transparent colourless glass, of a thickness of 

 one tenth of au inch. 



Nature or Experime:\ts. — In commencing my experiments I 

 at first employed cut shoots and detached leaves of various plants, 

 as has so generally been done by previous experimenters. In 

 every case they were weighed both before and after their expo- 

 sure for definite periods to dayliglit transmitted through the 



coloured glasses. DilTerent methods Avere adopted ; but the 

 most satisfactory for short periods, such as are indeed only pos- 

 sible with detached specimens, was to cut the shoot under distilled 

 water, inserting it in a small test-tube submerged at the same 

 time. This secured the cut end from exposure to the air. The 

 surface of the water in the tube was prevented from evaporaiing 

 by a few drops of oil forming a thin layer. The whole is easily 



