CANNON: Srupies iN Pranr Hysrips 159 
I would suggest that the transmission of a dioecious tendency is 
morphologically possible because the spores that give rise to the 
prothallia are pure, and, therefore, we conclude, tend to reproduce 
the same sort of prothallia as the ones from which they sprang in 
the preceding generation, and that the dtoecious habit is likewise 
only possible by a continuance of the same process through many 
generations. By referring to Fig. 3, the formula for a dioecious 
fern hybrid, it is seen first that the gametophytes are of pure de- 
scent, and further that the fern is dioecious, so I suggest that the 
spores A will produce prothallia like those from which the chroma- 
tin of their nuclei was derived and, also, that the spores @ will give 
rise to prothallia like a of the pure form, that is, in dioecious hybrid 
Jerns, or the pure races, the prothalha that arise from A or a will be 
male or female accordingly as the sexual ancestor of that spore was 
male or female. In other words, may not the purity of the spores 
be the groundwork upon which the selective processes play that 
bring about the dioecious from the monoecious condition, that 
deepen the dioecious tendency into a dioecious habit, and, finally, 
give rise to the morphological condition known as heterospory ? 
We may now return from this disgression to examine very 
briefly what happens when seed-plants of different sorts are crossed. 
We should note first that the formula commonly given for the 
hybrid fertilization and variation in seed-plants is somewhat mis- 
leading since it does not take account of the sexual generation, 
which we believe in the monohybrid seed-plants of the Pswm-type, 
if indeed not in all others, is always pure. Secondly, it is the 
heterosporous condition of the higher plants, in the place of the 
homosporous monoecious one of the fern mentioned above, that 
makes possible the formula(A + 4; A+a;a+4,;a+a) of the 
hybrid of the Piswm-type. Now we may ask, what morphologi- 
cal conditions are prerequisite for the development of this formula? 
It is clear in order that this end may be reached that the anthers 
must organize spores with A and spores with @ characters and 
likewise that the ovary must maturate functional spores with 4 
and with a characters. Now as a matter of fact in the higher 
plants three of the potential macrospores usually degenerate and 
one only germinates, so that in any ovule one character, and one 
sort of chromatin, are eliminated from the problem. Therefore 
