XCvi PROCEEDINGS OF THE ROYAL HORTICULTURAL SOCIETY. 
So far as can be ascertained, no other variety of Lily growing in the neighbour- 
hood has been touched, which would make it appear as though L. regale possesses 
some particular attraction for the pests. It is necessary to find some remedy, 
as the attacks of the ants are so persistent that a fine group of this beautiful 
Lily has been ruined within the course of a fortnight." It is probable that a 
disc of cardboard covered with Tanglefoot fixed round the stem some little 
distance from the ground would prevent the ants from reaching the flowers. 
Change of Colour in a White Currant. — Dr. Walter Dick sent a note calling 
attention to the complete reversion of a White Currant planted against a wall 
at Great Massingham, King's Lynn, to a red form. For two years prior to the 
present the plant bore white Currants, but this season all on the bush are red, 
and ripe ten days earlier than those on other bushes in the garden. 
Spencer-flowered Cupid Sweet Pea. — Mr. G. T. Dickson, Newtownards, sent 
a Cupid Sweet Pea with waved flowers of the Spencer type, which he had raised 
along Mendelian lines by crossing fin 191 3) a tall Spencer with the ordinary 
grandiflora type of Cupid Sweet Pea. The new form appeared in the second 
generation, and has bred true since. Mr. Dickson sent the following note : 
The great advantage of the formation of a chart based upon the Mendelian 
hypothesis by plant breeders before commencing hybridizing is very clearly 
demonstrated in the following experimenls. 
In 1912 I considered the possibility of what, to my mind, would be a valuable 
contribution to floriculture and at the same time provide an experiment of 
scientific interest. 
The experiment, I should further add, was planned on an economic, rather 
than an academic, basis; the objective I had in view was to put the modern 
"Spencer form " of the inflorescence of the tall Sweet Pea upon the Cupid or 
Dwarf type of Sweet Pea, eliminating frcm the latter variety the horticultural 
"out of date " Grandiflora inflorescence. 
I worked the problem out in theory thus : — 
Assuming one can cross successfully the tall Spencer with the Cupid Grandi- 
flora in 1913, the resultant hybrid in 1914 (Fi) generation would be tall, con- 
taining the gametes as undernoted. 
* Let t = Tall ; S = Spencer ; C = Cupid ; g = Grandiflora. 
t and g to be assumed Dominant factors. 
S and C to be assumed Recessive factors. 
The 1 91 3 cross would be : 
Tall Spencer x Cupid Grandiflora ; 
or, tS x Cg. 
resulting in (tSCg) ; Fi generation (1914). 
The 1 91 4 cross would be tSCg x tSCg. 
Thus the gametes in the above hybrid would be as the monoclinous flower plan 
on p. xcvii shows. 
This is a design of my own, and one which I think should simply illustrate 
the sexual unions as they could take place in an hermaphrodite flower. 
The foregoing, as I have already stated, was purely theoretical, but the 
experiment, covering the years 1913, 1914, 1915, 1916, and proved in 1917, has 
worked out entirely in agreement with the Mendelian hypothesis. 
When I commenced my experiment I was doubtful as to whether a sexual 
affinity existed in the two forms of Lathyrus I had chosen to experiment with, 
and indeed I may add that I had almost given up hope of effecting a cross between 
the two, for it was only after 130 operations that I established one single cross, 
With this all-essential success I assumed that the characters for tallness and 
grandiflora form would be dominant, and that the characters for dwarfness and 
Spencer form would be recessive, and that when I found the latter in a 
combined state amongst my cultures in the F2 generation I had bred the new 
form, viz. a Cupid Sweet Pea with Spencer flowers, which I had set out to 
accomplish. 
v My imaginary chart, which proved a reality, was not based on an unreasonable 
assumption, for although I was treading, to me, new scientific ground I had 
no reason to disbelieve that certain factors would be dominant and others 
recessive, as I had specified. 
The experiment suggests (a) that the old Grandiflora Cupid is probably a 
mutation from the tall form, and (b) that the Spencer form of inflorescence in the 
Sweet Pea has always been a latent factor. 
* I use the capital and small letters to enable a homogeneous and heterogeneous 
combination to be quickly observed. 
