WORK OF THE UNITED STATES DEPARTMENT, &c. 131 
The Common Orange and C. trifoliata are very distinct in character, 
and are somewhat difficult to hybridise. In my personal work I find 
that, even using the utmost care, only about one twentieth of the flowers 
pollinated seem to be affected by the cross, and only about one per cent. 
of the flowers cross-pollinated finally mature fruit. There is a further 
loss in the small percentage of germination, not more than half of the 
seeds germinating. We have secured quite a number of hybrids, 
however, where both the Orange and trifoliata were used as the seed- 
bearing parent. Some of these plainly show characteristics of both parents, 
and are doubtless true hybrids. Out of forty hybrids of the trifoliata 
crossed with pollen of the Sweet Orange, twenty-nine resemble the 
Trifoliate Orange in habit and foliage characters so far as can be observed, 
while eleven are clearly intermediate in these characters. These eleven 
intermediate plants are very similar to each other, and apparently derive 
certain characters from each parent. The leaves are trifoliolate but 
are much larger than those of normal Citrus trifoliata. The central 
leaflet has a tendency to enlarge, while the lateral leaflets remain 
about the normal size or in some cases are reduced in size. (Compare 
fig. 42, which gives a hybrid (772) between its two parent species, all 
three being seedlings of the same age.) 
The majority of the species of Citrus are polyembryonic, several 
embryos developing in one seed, and frequently giving several seedlings 
when germinated. This introduces an interesting complication into 
citrous hybridising. Strasburger * has shown that the egg cell proper is 
apparently fecundated in the normal way, and develops into a single 
embryo, and that the other embryos are developments from certain cells 
of the nucellus (the mother tissue), near the apex of the embryo sac, 
which become enlarged and divide, and finally push out into the embryo 
sac, forming what are termed adventive embryos. When fully 
developed in the mature. seed, these adventive embryos cannot be 
distinguished from the embryo developed in the normal way from ‘the 
fecundated egg cell. In many cases a seed from a hybrid fruit has 
yielded more than one seedling, and in several instances seedlings from 
the same seed are of totally different character, showing that one has 
been affected by the hybridisation while the others are like the mother 
parent. In such cases it seems evident that the intermediate seedling 
which shows the effect of the male parent is from the embryo, . resulting 
from the development of the fecundated egg cell, while the seedlings 
resembling the mother parent only, are developed from the adventive 
embryos, which, as explained above, arise directly from cells of the mother 
parent without any intervention of the male element. 
In the pot shown in fig. 48, No. 2, three seedlings are developing 
from a single seed of a Tangerine Orange which was crossed by Mr. 
Swingle with pollen of trifoliata. One of the seedlings has trifoliolate 
leaves, while the other two have unifoliolate leaves like the mother parent. 
There can be no possible doubt that these three seedlings come from the 
same seed, as this difference was plainly visible when the seedlings were 
about an inch high, and fearing that there might be an error I carefully 
*Strasburger, Dr, Eduard, “ Ueber Polyembryonie,” Jenaische Zeitschr. fiir 
Naturwissensch. XII. ' 
K 2 
