CHAP. VI. 
HYBRID PLANTS. 
287 
vents their reciprocal fertilisation, so does this obstacle, of 
whatever nature it may be, in general present an insuperable 
bar to the intercourse of different genera. All the stories 
that are current as to the intermixture of oranges and pome- 
granates, of roses and black currants, and the like, may, there- 
fore be set down to pure invention. 
It is, nevertheless, undoubtedly true that higeners^ that is to 
say mules between different genera, have in some few cases 
been artificially obtained. Kolreuter obtained such between 
Malvaceous plants ; Gaertner, between Daturas and Henbane 
and Tobacco ; Wiegman, between a Garden Bean and a Lentil ; 
and there are other well-attested cases. But all such produc- 
tions were as short-lived and sickly as they were monstrous. 
By far the best series of observations that has been instituted 
with a view to determine the laws of hybridism was that of 
Kolreuter, who, about the year 1775, commenced a set of 
experiments, which he continued to prosecute for twenty 
years, upon species of the genera Digitalis, Verbascum, Sola- 
num, Malva, Linum, Dianthus, and Mirabilis. It is upon 
those experiments, combined with the subsequent experience 
of others and my own observations, that the foregoing state- 
ment has been made. 
It has, nevertheless, been asserted by divers experienced 
cultivators of the present day, that the conclusions drawn 
from the experiments of Kolreuter have been too hasty ; and 
that, if they apply to the genera that were the special subject 
of the attention of that observer, they are by no means appli- 
cable to plants in general. It has been urged, in proof of this 
statement, that many different species of African Gladioli, of 
Pelargonium, of South American Amaryllis, of Crinum, of 
Triticum, &c., breed freely together, and that their seedlings 
are as fertile as themselves. 
I must confess that these instances are by no means such as 
to shake my confidence in the accuracy of the laws deduced 
from Kolreuter's experiments. In the first place, there is a 
degree of vagueness and looseness in the cases that are 
specified, which is particularly striking if compared with the 
precision with which Kiilreuter's experiments were conducted; 
secondly, in all the instances above mentioned, which, I believe, 
