1907] GATES POLLEN DEVELOPMENT AND MUTATION 109 
: 3. O. biennis and O. grandiflora, American members of the genus, 
f have been shown by MacDoveat to be mutating. 
_ 4. Among all the known types of hybrids, with their widely varying 
methods of transmission and of variation, none are known, so far as 
the writer is aware, which present a series of phenomena of variation 
resembling or comparable to those found in O. Lamarckiana and its 
mutants. 
5. Hybrids from crosses among the mutants of Oenothera follow 
the ordinary laws of hybridization; ‘e. g., O. lataXO. Lamarckiana is 
a Mendelian or dichotomous hybrid; and O. nanella XO. rubrinervis 
is a hybrid of intermediate type, which breeds true. 
6. If O. Lamarckiana has itself been hybridized, where did the 
forms originate with which, on this assumption, it was hybridized ? 
No scrap of evidence that they ever originated from any source 
except from O. Lamarckiana seed has been found. We must then 
assume that this is their only method of origin until the contrary is 
proved to be the case. 
Summary 5 
1. Failure of pollen development in O. /aia is not due to ingrowth 
of the tapetum to fill the loculus, as described by Pout, but 
to some other agency at work in the hybrid, the nature of which is 
not fully explained. 
2. Pollen development in O. /ata may proceed to the formation of 
tetrads, but degeneration of both mother cells and tapetum frequently 
begins in the resting stage or during the prophase of the first mitosis. 
3- Heterochromosomes arise in O. Jata in the prophase after 
synapsis. There may be one or two such bodies, which are formed as 
large rings by the cutting off of a portion of the spirem thread before 
the remainder breaks up into chromosomes. These bodies later are 
found in metaphase of the heterotypic mitosis, usually in the cytoplasm 
near the spindle. They do not divide, but become smaller and 
SAs this paper is passing through the press, conditions are being found in 
the pure O. Lamarckiana which will modify the interpretation of the bodies called 
heterochromosomes. The inference that O. Lamarckiana itself has the same number 
of chromosomes as the dominant O. Lamarckiana hybrid is also apparently not borne 
out by the facts. There seem to be variations in the number of chromosomes, which 
will require further careful study before an interpretation will be possible. 
