218 Margaret Hume. 
and the line of demarcation between the two components is there¬ 
fore a sharp one. In Cratcegornespilus and in some of the other 
Solarium graft hybrids, more than one of the outer layers of cells 
belong to the one component, and as the cells under the epidermis 
are not regularly arranged, it is impossible to say with absolute 
certainty to which component any one cell may belong, a state of 
affairs which makes these graft hybrids quite useless as a basis for 
argument upon the genesis of connecting threads. 
When the present investigation was begun, the writer was 
unaware that the existence of connecting threads had already been 
demonstrated between the two components ( Cytisus purpureus and 
Laburnum vulgare) of Cytisus Adami. This has been done by 
Buder (5, p. 214) and it only remains for the writer to confirm his 
observation of the connexion between the epidermis and the layer 
below it, by means of deep pits. 
The material of Cytisus Adami was obtained from the Royal 
Botanic Gardens, Kew, and the material of Solarium tubingense and 
Solatium Koelr enter ianum from the John Innes Horticultural 
Institute at Merton. 
The methods used were the ones generally employed for the 
demonstration of connecting threads. The material was fixed in a 
dilute iodine solution and mordaunted in a mixture of potassium 
iodide and uranium nitrate; it was then cut on the freezing micro¬ 
tome, stained in safranin and washed out with dilute glycerine and 
London blue. 
1. Cytisus Adami. To ensure that the tissue used was really that 
of Cytisus Adami and not material from any part of the graft hybrid, 
which might have reverted to one of the original components, a part 
of the flower was chosen, for it is in the flower that the characteristics 
of Cytisus Adami are most sharply distinguishable. The actual part 
of the flower used, was the thickened base of the vexillum, where 
it narrows down to the point of attachment. 
As has already been stated, the results obtained for Cytisus 
Adami were confirmatory of those obtained by Buder. Deep pits 
connect the cells of the epidermis with those of the underlying layer, 
but Buder (5, p. 214, text-fig. 1) depicts numerous delicate threads 
crossing the pit-closing membrane, which the present writer was 
unable to distinguish with certainty, though their presence is 
naturally to be suspected. The pits and their fillings do not differ 
in appearance from those which connect the cortical cells of Cytisus 
Adami with one another. 
