1918) BUCHHOLZ—PINUS 217 
Pseudolarix (30), and Tsuga (31) are reported as similar to Pinus; 
the embryo is said to consist eventually of 4 tiers of 4 cells each. 
In Sciadopitys LAWSON (25) found 8 free nuclei before organiza- 
tion into tiers takes place. This is very significant, for here we 
may have this extra free nuclear division result in more embryo 
initials, a thing which would bring about a greater display of cleav- 
age polyembryony than in Pinus. Judging from the figures of 
ARNOLDI (1), this conclusion seems justified, for the central group 
of cells shown in several of his figures is doubtless made up of many 
embryo initials from which the embryos are elongating. The 
writer believes that cleavage polyembryony is a very primitive 
feature, and it is therefore possible that the embryo of Sciadopitys 
is more primitive than that of Pinus. 
Little is known in regard to the rosette of other conifers. The 
work done on Picea (28), Abies (29), and Tsuga (31) does not include 
the stage showing the suspensor elongating. MryAKE and YAZzIN 
(30) have figured a stage in Pseudolarix with the suspensor elon- 
gated, which proves that a rosette group exists in this genus. It 
is not safe to conclude that a rosette exists in all forms in which the 
proembryo is organized in tiers like Pinus. 
In Pseudotsuga Lawson (24) reports a proembryo similar to 
Pinus, but does not show which tier of cells elongates, or whether 
a rosette exists. He applies the term ‘‘rosette”’ quite generally to 
the upper tier of free nuclei where no rosette cell group exists. 
Likewise, CoKER, in his work on Podocarpus (6) and Taxodium (7), 
uses the term “‘rosette”’ to designate the group of free nuclei above 
the suspensor. While these investigators apply the term “‘rosette”’ 
here, it is evident from a comparison of the figures that a rosette 
homologous to that of Pinus does not exist in Podocarpus, Taxo- 
dium, or Cryptomeria. The term “rosette,” as first used by 
MrrBEL and Spacu (27), applies to an unelongated tier of com- 
pletely walled cells. Lanp (20) showed that it is likewise the 
_ uppermost tier of completely walled cells that elongates in Thuja. 
The absence of a group of rosette cells and of rosette embryos is a 
more advanced character, found only in the more recent conifers. 
SAXTON (34) has described the embryo of Actinostrobus, which 
repeats the proembryo of Sequoia in completely filling the egg 
