1907] GATES—POLLEN DEVELOPMENT AND MUTATION 103 
In some of the papers sufficient attention has not been given to a 
comparison of the adult characters of the hybrid with those of the 
parents, 7. e., whether the hybrid is a mosaic, Mendelian, an inter- 
mediate, or goneoclinic to one parent, constant and invariable or 
breaking up into different types, etc. It would be particularly valu- 
able to know in every case to what particular type the hybrid whose 
germ cells are studied conforms, and what variations if any occur. 
THE POSSIBLE CAUSES OF STERILITY IN HYBRIDS 
That hybrids are frequently sterile has of course long been known 
and commented upon. The recent morphological and cytological 
papers already cited have not reached an explanation of its cause, 
though various suggestions have been made. Some of the causes 
suggested may be classified into several general categories which are 
not, however, necessarily mutually exclusive: (1) lack of “nutrition” 
in the parts affected, causing, e.g., failure in development of the 
tapetum or low staining capacity of the chromatin in the germ cells; 
(2) the influence of long culture (whatever that may be); (3) irregu- 
larities of development in the germ cells, appearing during synapsis 
or the reduction divisions and due to some lack of harmony (variously 
expressed) between the idioplasms; (4) some more deep-seated 
phenomenon affecting the physiology of the whole plant. This is not 
an attempt to classify all the possible causes of sterility suggested, but 
it will be seen that the factors are stated from widely differing points 
of view and with varying amounts of definiteness. Some of them 
might be viewed as effects rather than as causes. 
In O. lata, if the tapetal cells always showed a degeneration before 
the pollen mother cells, we might conclude that the failure of the 
tapetum to function (in the secretion of nutriment for the pollen 
grains, which is generally conceded to be its chief function) was the 
immediate cause of the abortion of the pollen. But this is not always 
the case; for degeneration in the pollen mother cells may appear as 
early as in the tapetal cells or earlier; while GREGorY states that in 
the hybrids of Lathyrus odoratus the “somatic cells” are normal 
throughout. Guyer (10) suggested in 1900 that incompatibility 
of the chromosomes or plasms, as shown in the prophases of the first 
reduction division, is the source of the subsequent irregularities in 
