1899] DEVELOPMENT OF THE MICROSPORANGIUM 335 
aswell as extensive. Careful examination shows that it is not 
acontinuous ribbon at this. period, but rather an anastomosing 
aetwork in which large granules of chromatin of unequal sizes are 
imbedded, giving it thus a more or less knotty appearance. 
Ata period quite early in the development of the anther the 
chromatin network contracts into a dense and ultimately spher- 
_ttalmass, which in most cases is in contact with the nuclear 
fembrane, but may also lie free in the nuclear cavity (fig. 8). 
The ball becomes so dense that it is ordinarily impossible, even 
with the weakest stains, to distinguish the separate threads of 
Which it is composed, except at the periphery where they pro- 
ket into the nuclear cavity outside. This condition is synapsis 
|Convallaria. The nucleolus does not lie inside the mass and 
adjacent to the membrane, as has been claimed for some other 
Plants. On the contrary, it always lies outside so far as could 
te determined, but usually in contact with the mass and at the 
‘ide away from the wall more often than toward it. 
Judging from the number of preparations obtained, the 
. tucleus must remain in the synapsis condition for a considerable 
ime, The first indication of the return from synapsis is found 
| the gradual Separation of the outermost linin threads, and 
Tally the whole mass becomes more open (fig.9). Buta vey 
Mange thing now occurs. At the moment when the chromatic 
2 ad is Spreading out, between its meshes and in the cavity 
de are to be seen large granules, or more properly speaking 
pe of various sizes which are decidedly chromatic, in fact 
: o “xactly like chromatin. The nucleolus meanwhile still 
ous intact, Where these bodies come from or where they 
Mt, at th: determined. They were always eins ber 
: Meleolus ig Stage. The masses were often as po 5 as 
c. a more often smaller. It might be ee ual 
Wy the Se the nucleolus becomes fragmented either n at 
Sport pate of the reagents, but the preparations 
Mote aa In very weakly stained nee - caG 
si. €olus with its central vacuole coul - ap 
: in the same nucleus the expelled masses were stat 
