740 Henslow. — The Origin of Monocotyledons from Dicotyledons , 
cell is formed a long tube which grows towards the chalazal end of the 
ovule, forming a passage by the absorption of the nucellus. It thus plays 
a nutritive part comparable to that of the antipodal cells in many genera. 
A similar formation of endosperm has been described in Sagittaria by 
Rendle. 1 
The details should satisfy the reader that they are all the result of an 
aquatic habit of life, thus bringing all Monocotyledons into close connexion 
with aquatic Dicotyledons. 
17. Speculations on the Arrest of one Cotyledon. 
Perhaps the most difficult question to answer is: What were the exact 
conditions under which one cotyledon was arrested ? Such arrest occurs in 
plants associated more or less with water now, as Ranunculus Ficaria and 
Pinguicula , hygrophytes ; or which were probably ancestrally aquatic, as all 
Monocotyledons. But the reduction of the embryo may take place also 
under other conditions of life, as in parasitism, by drought, &c. 
We have seen above that it was clearly proved that the arrest of the 
intercostal tissue in a submerged leaf is actually due to the weakening effect 
of water by saturating the protoplasm, so that it is unable to develop a 
complete blade. From this we may confidently generalize upon the effect 
of an excess of water upon all parts of aquatic plants. 
When we try to connect water with a reduced size of the two cotyledons 
in the seed or with only one, or again with their total arrest, we are con¬ 
fronted with the fact that existing aquatic Dicotyledons have, as a rule, 
two cotyledons ; but as such aquatic genera or species mostly belong to 
present terrestrial orders, they are of comparatively modern origin, and we 
can only assume that heredity has overcome any deteriorating effect sufficient 
to totally arrest one cotyledon. Still, the size of the cotyledons is not in¬ 
frequently greatly reduced when in the seed; indeed, they may be arrested 
in the pro-embryonic condition, so that we seem to see degrees of degrada¬ 
tion. Thus:— 
1. The embryo may be relatively minute as in Ranunculaceae, in 
Dionaea , Drosera , Gunner a , and Eriocatdon , more or less aquatic plants. 
On the other hand the xerophytic Holly has a very small embryo embedded 
in a mass of albumen. 
2. In some seeds the cotyledons are small and arrested as compared 
with the size of the radicle, e.g. Pinguicula , genera of Halorageae, Parnassia. 
3. The embryo may remain in the seed arrested at the pro-embryonic 
condition of an undifferentiated globular mass of cells, as in Ranuncidus 
Ficaria , several species of Anemone , Utricidaria^ &c. 
4. In the next place there may be an inequality in the size of the 
embryo, as in Nymphaea , &c. 
1 See also Class, of Flowering Plants, i, 163 (Rendle) ; and Bot. Gaz., xxiii, 6, 260. 
