AND ON POLYEMBRYONY IN THE HIGHER PLANTS. LO 
the place where the embryo is subsequently formed. It is not 
even proved that impregnation in the higher plants is always 
effected by a pollen-tube. I myself have related cases where I 
had never observed the pollen-tube in young impregnated ovules ; 
and indeed in cases which I have so frequently examined, that 
I am actually inclined to doubt that impregnation is there 
effected by a penetrating pollen-tube. But the discoveries of 
MM. de Mirbel and Spach on the ovules of the fruit of Mais, 
are still more striking ; for at all events impregnation cannot, in 
this instance, be effected by a penetrating pollen-tube, but is 
perhaps to be explained by a dynamic action, which the fovilla 
may exert on the primitive utricle, which is transformed into the 
embryo. 
Just as this little work was going to press, a memoir by M. 
Bernhardi appeared*, entitled, “ On the formation of seed with- 
out previous impregnation,” which also treats of the subject 
here under consideration. M. Bernhardi believes that an adjust- 
ment of the different views which at present prevail on the 
sexes of plants, depends principally on three points: 1. on con- 
tinued direct observations of the changes that take place in the 
vegetable ovula after the pollen-dust has been shed [nach der 
Bestiubung]; 2. on the determination of several phanomena 
which the production of hybrids presents; and 3. on the con- 
firmation or refutation of observations, according to which nu- 
merous plants are capable, under favourable circumstances, of 
producing germinating seed even without previous impregna- 
tion. Now the recent memoir of M. Bernhardi treats especially 
of this last point; and it is well known that at various times new 
voices have continually been raised in favour of it. In the first 
place, M. Bernhardi enumerates several cases of animal propa- 
gation as to which a conviction has been come to that no pre- 
vious impregnation had taken place; and if anything of this 
kind can occur in animals (for the incorrectness of those state- 
ments has not yet been proved), it would be conceivable in 
plants also. In relation to this the celebrated experiments of 
Spallanzani are cited, and all the observations of others referred 
to, who have sometimes disputed the results of those experiments, 
sometimes confirmed them, but more especially the observa- 
tions on hemp plants, which have recently (1828) been re- 
peated by Girou de Buzareingues, and still earlier also (1811- 
* Algemeine Gartenxeitung von 1839, vom 12 October. 
Cc 2 
