CANNON: Srupies IN PLantr Hysrips 147 
tiberhaupt keine Bastarde (Papilionaceen, Coniferen, Urticaceen, 
Convolvulaceen u. s. w.). Dasselbe abweichende Verhalten findet 
sich unter verwandten Gattungen und Arten vor. Weinreben, 
Weiden, Dianthus-Arten sind leicht, Sz/ene-Arten schwer, die von 
Nicotiana, Verbascum, Geum leicht, die Arten von Solanum, 
Linaria, Potentilla dagegen schwer unter einander zu bastardiren. 
Eine Hybridisirung von nahe verwandten Arten will oft nicht gelin- 
gen (z. B. Apfel- und Birnbaum), dagegen lassen sich Pfirsich mit 
Mandel, ja sogar Species der verschiedenen Gattungen Lychwis und 
Stlene, Rhododendron und Azalea, Aegilops und Triticum kreuzen, 
je nach ihrer ‘ sexuellen A ffinitat.’ ” 
The suggestion of Strasburger that crosses between mosses 
and ferns cannot take place because of the difference in the chem- 
ical nature of the fluids attractive to the spermatozoids of the two 
forms, may be a fruitful one when applied to other plants that are 
difficult or impossible to cross. That is, the substance that attracts 
spermatozoids, or the pollen-tube, may vary with the plant, and 
instead of attracting may even repel them. So far as I know the 
course of the pollen-tube in plants that are nearly enough related 
for crossing, but which refuse to be crossed, has not been traced, 
and it is possible that here may be found one reason why certain 
reciprocal crosses will not take place, why nearly related plants 
may sometimes not be crossed, and also why one sort of pollen is 
‘‘prepotent’’ over another sort. That the pollen-tube responds to 
such stimulus in the course of its wanderings most sensitively is 
well shown by Lloyd’s account * of its behavior in Diodia and 
Richardsonia. There may be other factors, as Vernon + has found 
that the relative maturity of the sexual elements in some echino- 
derms is of great importance when the number of successful 
crosses, and even the effect of each parent on the offspring, are 
concerned. 
While it is perfectly true that only nearly related forms as a 
rule may be crossed, there are many instances where hybrids 
have been made whose parents were generically and even more 
remotely separated. These, as was mentioned above, are usually 
wholly sterile. A few such hybrids may be mentioned here. 
* Lloyd. Mem. Torrey Club, 8: 88. 1902. oes 
+ Vernon, Hybrid Echinoid Larvae. Proc. Roy. Soc. 63: 229. 1099. 
