334 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
But meanwhile my colleague J. Vallerand, sen., had crossed the 
Sinningia-Gloxinia, of a lilac-white colour, dotted all over with dark 
violet, with Gloxinia erecta. As the result of this hybridisation, he 
obtained some very curious and new things: plants with very finely 
spotted and marked flowers. It was a veritable revolution! During 
this time I acquired from M. Eugéne Vallerand a series of Gloxinias 
(Ligeria), in connection with which it will be necessary to say something 
These Gloxinias were of a very special type, and I believe were obtained 
from an able cultivator named Rossiaud. Very distinct in habit, these 
plants had rounded thick leaves, which, as well as the stems, were of a 
velvety nature; the buds were plump, short, and strongly petiolate; the 
flowers were large, of fine substance and very wide open; but they had a 
very narrow tube, and the colours, which were not very diverse, ranged 
from dark blue to cherry red, passing through bright rose or a white 
ground marked with blue dots. 
They were, in short, very distinct Gloxinias, having nothing in 
common with those we possessed before ; nor with those of the elder Valle- 
rand which had been crossed with Simningia. They were afterwards given 
the name of ‘ Crassifolia.”. Where did these types come from, and what 
was their origin? For they bear so little resemblance to the male parent 
from Brazil, that it was easy to discriminate them at first sight. To fix 
this type, improve it, give it the colours wanting, and above all render it 
floriferous and less stiff—that was my task. I devoted ten years to it, 
from 1869 to 1879/1880, and during that period I have been enabled to 
obtain not only a great diversity in colour, but also pure whites, and 
moreover exceedingly floriferous plants. Under the influence of cultiva- 
tion and continual selection I saw the type transformed, produce 
numerous buds, compact and abundant flowers, and assume a totally 
different aspect. 
The origin of these Gloxinia crassifolia is indeed a mystery. Is it a 
type come spontaneously, and which has thus produced itself alone? I 
do not think so, for the first Gloxinia (Ligeria) crassifolia had drooping 
flowers, and their general appearance resembled that of the Gloxinia 
pyranidalis. Did the intervention of the latter take place at the 
opportune moment, and with what variety ? Or was it Rossiaud who, 
wishing to fertilise the Gloxinia pyramidalis, did so with the Gloxinia 
(Ligeria) obtained direct from Brazil or already improved on by him ? 
At any rate, one thing is certain, and that is that the Gloxinia crassifolia, 
although improved, have never allowed themselves to be fertilised by 
the Sinningias or their descendants. I have tried this operation several 
hundreds of times, but without success, and I leave it to others to tell us 
when they have proved successful. 
The type crassifolia has at no time had dotted flowers; it is fixed and 
unchangeable; the colours have been improved, and, as I have already 
remarked, I have been able by dint of patience and time to transform it 
in such a manner that it is now quite as free-flowering and as rich in 
colours as the type called ‘ Vallerand’s Type.’ How has this transfor- 
mation been effected? It is the result of repeated fertilisations, it is 
true; but one must also take into account the improvements which were 
produced as I proceeded with the fertilisations. 
