1915] BURLINGAME—ARAUCARIA BRASILIENSIS 13 
fig. 40 or 41. I have found no seed with a second embryo large 
enough to be seen with the naked eye. 
Ever since STRASBURGER’S account (23) of the cap of the pro- 
embryo of Araucaria brasiliensis, it has excited comment on its 
apparent specialization. It has been spoken of as a protecting 
cap (7, 20, 23), but no evidence has been adduced to show that 
protection is at all necessary. The caps are not made of specially 
strong cells, nor do they show any effects of abrasion, which might 
reasonably be expected if they were of use as a protection. Neither 
do the cells of the endosperm surrounding the caps appear to have 
been crushed and thrust out of the way. It seems much more 
likely that the cells of the cap secrete a digestive enzyme. An 
inspection of figs. 34-37 will make this evident. Very few cells 
of the endosperm show any distortion from crushing, while prac- 
tically all of them show the action of some corrosive agent on 
their contents or even in some cases on the walls themselves. 
I have seen no evidence, however, for thinking that the secretion 
of enzymes is limited exclusively to the cap. In fig. 34 it will be 
seen that the region of greatest cell destruction is around the 
embryonic region and not directly in front of the advancing tip. 
It is clear from fig. 37 that the cavity in which the proembryos lie 
continues to enlarge around the suspensors long after the cap has 
passed by. To put the argument in another way, the cavity 
should be cylindrical if solution occurs only around the cap, whereas 
the cavity is actually shaped like a wide-mouthed cone, showing 
that solution has gone on all over its surface and not merely at the 
apex of the cone. It is, of course, possible that this might be true 
and still all of the enzymes be formed in the cap, but excreted in 
such abundance that they fill the entire cavity with a solution of 
equal strength. I suspect that the matter comes to about this. 
The cap looks like a highly specialized structure and should in 
consequence have a specialized function. The proof that it does 
actually have a special function has not yet been adduced. 
Sometime in June or July the primary suspensors have reached 
their limit of elongation. Then begins the third and final stage 
of development of the proembryo. The activities of this stage are 
limited to the embryonic group of cells. A rapid multiplication 
