258 DUPLICATION OF GENERATIVE NUCLEI BY MEANS OF 
they are still in an unripe condition. The „Weekblad voor Bloembollen- 
cultuur 21St® Jaargang No. 78. 28 Maart 1911, relates that the old 
breeders treated their bulbs in former times, as follows: „We used to 
cover our bulbs, which usually were taken up, while still green, by a 
layer of soil, left them there for a fortnight or so, and subsequently 
brought them to the bulbhouses”. What was done in 1911 is thus des- 
cribed: , All bulbs, with the leaves dead or alive, are taken up towards 
the middle of June and afterwards exposed, during a certain time, 
to a temperature of 75° tot 80° Fahr.” 
Up to 1890 the way in which and the time at which the bulbs were 
harvested, hardly underwent any change. The leaves, green, yellow, 
or dry were all simultaneously shoveled off. Before Petrus and Paulus 
(the 29th of June) all bulbs had to be covered by a layer of sand and 
were thus left in the pathways between the flower beds. Artificial heat 
was at that time not yet applied. The drying of the bulbs took place 
in the sun and afterwards in the bulbhouse. Between 1890 and 1900 
the application of artificial dry heat began to be applied in the bulb- 
houses and soon came into general use, a custom which has had impor- 
tant consequences for the cultivation; a point which I shall not further 
discuss at the present moment. 
It seems to me, that it has sufficiently been shown, that, during some 
centuries already, the varieties of Hyacinthus orientalis have been exposed 
to external conditions, which, as we have seen, favored the formation of du- 
plication of generative nuclei. 
I am furthermore of opinon, that we may conclude from this: 
1°. unconsciously cultivators of hyacinths have pollinated their plants 
with the abnormal fertile pollengrains, mentioned, while such has 
also been done by insects; 
2°. this in all probability, has been the cause of the origin of triploid 
descendants; 
3°. as these triploid plants were often characterised by an unusual size 
and vigour, they were, provided they were satisfactory in other res- 
pects also, preferred to the smaller ones; 
4°, in the beginning, crossing will have occurred between the many di- 
ploid and the few triploid hybrids. Because, as has been shown, the 
reduction division of the triploid varieties proceeds irregularly, the- 
se crosses gave a progeny with a diverse number of chromosomes in 
the somatic cells; ar 
