322 TRANSACTIONS AND PROCEEDINGS OF THE [Sess. lxiv. 



open dish for at least two mouths during winter without 

 in the least affecting their power of germination, though if 

 this process has commenced in the smallest degree, even a 

 very short period of dryness is sufficient to kill them. 



That the layer of mud under which the buds require 

 to be buried acts merely as a screen, cutting off the light 

 rays, may safeh'' be concluded from the fact that when 

 buds in a vessel of water are protected in any other way 

 from the action of direct illumination, as, for example, by 

 enclosing the vessel in a metallic case, germination does 

 not take place until they are again exposed to light. 



Intense illumination, such as bright sunlight, is not 

 necessary for germination, as this process goes on freely, 

 though somewhat more slowly, in diffused light, such 

 as that supplied by a north window. On the other 

 hand, the minimum degree of illumination below which 

 germination will not take place is far removed from 

 absolute darkness, and buds will remain unchanged for 

 long periods if merely shaded from direct light, or kept at 

 a considerable distance from a window. 



With the. intention of approximately determining what 

 particular rays exert the greatest influence in bringing 

 about germination, use was made of a series of coloured 

 screens, composed of aqueous solutions of various inorganic 

 salts. The following were found most suitable, and were 

 all carefully tested spectroscopically : — 



1. A solution of ferric thiocyanate, of such a strength 



that only light of wave lengths between 670yu./A 

 and V 1 5 fjifx, was able to pass through a thickness 

 of 20 mm. It thus transmitted only the deep red. 



2. A dilute solution of the same salt, which transmitted 



all the rays between 570 /n/u. and 715 fifi, i.e. most 

 of the red, all the orange, and part of the yellow. 



3. A .saturated solution of potassium bichromate, which 



allowed all the rays between 522 /jl/jl and 715/i/x, — 

 i.e. nearly all tlie red, the orange, yellow, and a 

 trace of green — to pass. 



4. A saturated solution of nickel sulpliate, which 



transmitted rays between 422/i/Lt and Gl 6 fjb/jb, 

 i.e. the blue-violet, bright blue, green, yellow, and a 

 little orange. 



