Edith Chick on Torreya Myristica. 85 
morphological upper surfaces throughout the greater part of the 
region within the seed. 
In PI. viii. fig.. 1, A represents one cotyledon, which is about half 
the length of the other. The tip of A is only free from the other 
cotyledon during the last { mm. of its length ; it also shows a ten¬ 
dency to lobe in its free region and the vascular bundle divides. 
The other cotyledon B begins to form a lobe C, 6 mm. behind 
its apex; the lobing in this case is complete, one branch of the 
the bundle passing to C. 
In the third seedling, (PI. vii., Fig. 3 and PI. viii., fig. 4), the 
apices of the cotyledons were bent into a sickle shape and the 
petioles, free for about half their length, became fused before this 
region was reached. Here again one cotyledon was much shorter 
than the other, and both showed a tendency to lobing in the 
behaviour of the vascular bundle, which widened and seemed about 
to separate into two portions in each. 
Strasburger 1 , in his work on the Coniferae and Gnetaceae, 
mentions this tendency for the cotyledons to be fused and lobed in 
Cycads and in Ginkgo , and it is especially interesting to find the 
same characters in two out of three specimens of Torreya, another 
primitive genus. 
In Ginkgo biloba the cotyledons remain underground in the 
endosperm and only the basal parts project from the seed-shell. 
They are not fully fused at their apex, but they adhere so closely 
together that they cannot be separated without tearing the tissue ; 
and what especially recalls the Cycads is the fact that the extreme 
tips of the cotyledons are swollen and show a feeble indentation, 
just as Schacht 3 figured for Zainia spiralis. In Torreya Myristica 
the lobing in one instance is much more marked than in Ginkgo , 
and fusion of the cotyledons was found in all three seedlings. 
It is hardly possible to say what may be the importance of 
this tendency to lobe. The cotyledons are no doubt influenced by 
space relations within the seed, and the irregular form of the 
endosperm may cause the variation in shape and lobing of the 
cotyledons in this case ; for in other cases of ruminated endosperm, 
as among the Palms and in the nutmeg, the lobing of cotyledons 
is frequent. 3 On the other hand in the case of Anona , where there 
is a ruminated endosperm, there is no trace of lobing in the 
cotyledons. 
1 Die Coniferen unci die Gnetaceen, p. 320. 
2 vSchaclit, Der Baum, p. 53. 
* Tschircli, Ann. du Jard. Bot. de Buitenzorg, Vol. IX., p. 165, 1890. 
