382 
Notes on Recent Literature. 
may be regarded as forms in which the gametophyte, while retaining 
a relatively primitive number of nuclei, has evolved along three 
divergent lines. On the other hand, those who regard them 
as specialised from the ordinary form explain the presence of the 
sixteen nuclei in the following way: It must be remembered that 
in most cases the embryo-sac is the one surviving and developing 
member of a row of four spores, so that five successive divisions 
(two to form the spores and three the eight-nucleate embryo-sac 
within one of these spores) intervene between megaspore mother¬ 
cell and egg. Undoubtedly this is more primitive than the state of 
affairs found in such a form as Lilium, where no row of spores is 
formed, so that only three divisions intervene between mother-cell 
and egg. Now, in Peperomia, Gunnera, and the Penteacete the early 
stages have been carefully studied to determine this point, and in 
each it is found that the mother-cell forms no row of spores, such as 
would be expected in a relatively primitive form, so that only the 
four divisions which form the sixteen nuclei intervene between 
megaspore mother-cell and completed embryo-sac. This fact, 
in conjunction with the tetrahedral arrangement of the nuclei 
formed by the first two divisions (an arrangement so common among 
spores) has been thought by some botanists 1 to indicate that in these 
cases the embryo-sac includes four megaspore nuclei, each of which 
divides twice. It is significant in this connection to note that genera 
allied to Peperomia (Piper 1 and Heckeria 2 ) and to Gunnera ( Hippuris 3 ) 
form a row of four spores, from one of which a normal embryo-sac 
is produced. 
This conception of the production of one embryo-sac from four 
megaspores, which germinate simultaneously, has even been 
extended to all cases in which—to use the current phrase—the 
mother-cell becomes the embryo-sac directly, the formation of spores 
being omitted. 4 Coulter is the chief exponent of this view ; he holds 
that the nuclei formed by the reduction division in the mother-cell 
are always to be regarded as the nuclei of four spores, whatever 
may be their subsequent history—so that in a case like L ilium, 
where there are only three divisions from mother-cell to egg, the 
gametophyte consists of only one generation of free nuclei. 
This view has naturally excited a good deal of controversy; 
the arguments that have been advanced for and against it may be 
summarised under four heads:— 
1. Five, four or three successive nuclear divisions intervene 
between mother-cell and egg in the Angiosperms. The first two 
divisions in the mother-cell (heterotype and homotype) show the 
same characters throughout the plant-kingdom, and always result in 
the formation of four spore-nuclei containing the gametophytic 
number of chromosomes. The further divisions of each of these 
nuclei are of the ordinary type, and belong to the gametophytic 
generation, whether this is formed from one spore only, or is a 
compound structure formed by the simultaneous germination of 
several spores within the mother-cell. 
Against this view it is argued that there is a general tendency 
in vascular plants to reduce their sporogenous tissue. “ Since we 
can trace the reduction of these divisions until among angiosperms 
1 Coulter, 190S; Brown, 1908, etc. 2 Johnson, 1902. 3 Fischer, 1880. 
< Lloyd, 1902, Davis, 1905, p. 471 472 ; Pace, 1908, 1909 ; Coulter, 
1908. 
