452 Messrs Gardiner and Hill , The Histology 
accomplished, the cotyledons are withdrawn from the depleted 
seed and unfold in the air, where they perform the functions of 
the ordinary green leaves 1 . 
In making a comparison between Tamus communis and 
Galium Tricorne it is interesting to note the great difference 
between the behaviour of the two seeds during germination. The 
composition of the cell walls appears also to be different in the 
two cases. In Galium , judging from microchemical reactions, it 
approaches more nearly to a pecto-cellulose, but in Tamus the 
walls appear to be composed of the so-called “ reserve-cellulose,” 
very similar to that described by Gruss 2 in the endosperm of 
Phoenix sylvestris. 
It is not unlikely that the character of the enzyme action 
during the germination of these two seeds has some relation to the 
differences in the composition of their cell walls. 
In the case of Tamus the mode of attack of the enzyme, which 
apparently is secreted by the cotyledon, is, with reference to the 
embryo, centrifugal , whilst in Galium , on the other hand, it is 
centripetal ; in this latter case it is the cells of the endosperm 
which appear to contain the enzyme, or perhaps rather a zymogen 
which gives rise to an enzyme. It is possible that the formation 
of the enzyme or zymogen may be determined by a stimulus 
chemical or otherwise proceeding from the embryo. 
On germination of the seeds the enzyme is liberated in the 
cells and commences to attack the walls and work towards the 
embryo in a centripetal manner. In both cases it was noticed 
that although in the first instance the walls were frequently 
entered by a ferment along the course of a thread, their subsequent 
disorganization took place without any definite relation to the con- 
necting threads in the cell wall. 
Strasburger 3 , who has recently examined the endosperm of 
Tamus in connection with his work on “connecting threads,” 
forms a somewhat different conclusion. He confirms the results 
which had been previously obtained 4 , but he considers that the 
chief function of the threads is to serve as the passage ways by 
which the enzymes enter and travel through the walls and effect 
their solution during germination. Further, he attempts to prove 
that it is the threads which traverse the thickness of the wall (the 
“ solitary threads ” of Kohl 5 ) rather than those found in the pit- 
] Cf. Lubbock, loc. cit. 
2 Gruss, “Studien fiber Reserve-cellulose,” Bot. Cent., 70, 1897, p. 242. 
3 Strasburger, ‘ k Ueber Plasmaverbindungen pflanzlichen Zellen,” Jahr. f. IFiss. 
Bot., Bd. xxxvi. Heft 8, pp. 535 — 588. 
4 Gardiner, Proc. Boy. Soc. 1897. 
5 Kohl, Ber. d. Deut. Bot. Ges. 1900. Dimorphic threads had been previously 
described in the paper in the Phil. Trans. Roy. Soc. 1883, in the endosperm of 
Bentinclda, Howea, Lodoicea and Phytelephas, with figures. Cf. also Gardiner, 
Proc. Roy. Soc. 1897, p. 104. 
