a a 
1900] DEVELOPMENT OF THE POLLEN GRAIN 95 
nucleus passed into the pollen tube while still inclosed within 
its cell membrane ; and several preparations showing stages fol- 
lowing division indicate that a membrane surrounds the two 
nuclei for some time (fig. 74). From the same figure it will be 
seen that the chromosomes of the daughter nuclei have not 
formed a dispirem, but remain for some time in the segmented 
form. What the form of these nuclei may be at a later stage, I 
have not been able to determine. 
NOTES ON MATERIALS AND ON METHODS. 
Collections.— In Symplocarpus I failed to secure any material 
earlier than October 1, and at this time the definitive microsporic 
archesporium was formed, and the nuclei of the pollen mother 
cells were well in the resting condition. The resting condition 
was found as late as January 1; but much of the material col- 
lected during late December and early January was in synapsis. 
Division of the pollen mother-cells occurred abundantly during 
the first warm days of February. Although a true spike, the 
order of maturity of the anthers in Symplocarpus is acropetal, 
in spite of the fact that the receptivity of the stigmas is often 
Synchronous. It is essential in collecting material to make an 
€xamination of a median anther to ascertain the general stage of 
development, although in exceptional cases there may be great 
Variation in a single spike. It is not uncommon to find all stages 
. from an early spirem to a late stage of division in the pollen - 
mother-cells. However, the development seems to go somewhat 
in waves, Usually all are in synapsis at the same time; then 
there is a rest after the second division of the pollen mother- 
cells, and there is no actual division of the microspore nucleus 
_ in any anther until at least the spirem stage of this division has 
been reached by all. 
For the best penetration of the fixative, it is necessary to 
Open the anthers, and this is readily effected wholesale by shav- 
ing off a little of the outer portion of the spike. The stamens may 
be again cut off next to the surface of the axis, and readily sepa- 
rated from the parts of the perianth. After fixing twenty-four 
