NOTES ON MELILOTUS ALBA 
421 
of an individual nucleus (25, Plate XIV). In many instances 
the formation of cross walls was initiated but suppressed at 
various stages. In some specimens the rudiment of a cross wall 
formed only a narrow trabecula (26, Plate XIV), while in 
others the trabecula extended almost across the cell width (28, 
Plate XIV). Several cases were found where septation had actu- 
ally taken place resulting in the formation of two grains from 
one mother cell (29, Plate XIV), and a few cases showed a di- 
vision wall cutting off one member of a tetrad (30, Plate XIV). 
Another case of occasional occurrence was that of mother cells 
which appeared to have passed through division normally but 
whose resulting pollen grains remained clinging tenaciously to- 
gether. To summarize, there were found all stages of gradation 
between pollen formed from mother cells without cleavage to 
pollen formed normally through mitosis and cleavage. 
DISCUSSION. 
There is a possibility that the giant grains were produced by 
hybrid plants. According to Jeffrey, giant grains have been of 
frequent occurrence in hybrid pollens examined in his labora- 
tory. Professor Jeffrey has no data on the nature and formation 
of these giant microspores. 
The collection of Melilotus flowers, which showed the aberrant- 
structures was taken from a number of plants and as the giant 
grains were found all through the collection it is probable that 
they were produced from a number of plants. If such were the 
case and if we are to assume that the large grains are in some 
way related to hybridism, then there must have been a number 
of hybrid plants growing in the same small area from which the 
collection was taken. Taking all the facts into consideration it 
does not seem probable that the giant grains can be safely as- 
cribed to hybridism. 
Cases somewhat similar to that found in Melilotus have been 
known for a long time in some of the Cyperaceae. Elving (1), 
Wille (8), and Strasburger (7) studied the pollen development 
in Eleocharis and later Juel (3) repeated the work very care- 
fully in Car ex acuta. In these forms the pollen mother cell 
passes through the tetrad divisions but three of the nuclei de- 
generate while the fourth persists and later divides into tube 
and generative nuclei. In these forms, however, the production 
of one pollen grain by a mother cell is the normal procedure. 
