210 Nichols, A morphological study of Juniperus communis var. depressa. 
protoplasin of tlie mötlier cell into fonr parts takes place, and the 
cytoplasm which snrrounds each of tlie nuclei encloses itself witb 
a thin cell membrane, bnt that no tbick walls, like tbose described 
by Miss Ferguson, are formed. From the following account, how- 
ever, it will be seen tbat tbe conditions in var. depressa are 
remarkably similar to tbose in Pinus. Tbat tbese features bave 
not been more frequently observed may be accounted for by tbe 
fact already pointed out by Miss Ferguson tbat tbe tbickened 
walls usually react very poorly to stains. 
Tbe structure of tbe microspore motber cell at tbe completion 
of tbe bomotypic division is sbown by fig. 35. Tbe fonr nuclei of 
tbe tetrad are connected witb one anotber in all directions by 
kinoplasmic fibers. Some of tbese represent connecting übers wbicb 
were in tbe spindle of tbe bomotypic division, and wbicb bave not 
yet disappeared. Otbers, bowever, arise de novo from tbe cyto¬ 
plasm, and to all appearances tbese are exactly like tbe true spindle 
fibers. Tbe übers soon become tbickened at tbe equatorial planes, 
giving rise to tbe six cell plates wbicb separate tbe fonr nuclei 
(fig. 36). Anotber important cbange, bowever, bas already taken 
place in tbe motber cell. Düring tbe anapbase of tbe bomotypic 
division tbere begins to appear in tbe peripberal region a trans¬ 
parent, bomogeneous layer wbicb stains bluisb witb gentian violet 
(ügs. 33—35). Tbis becomes tbicker, and wben seen in section 
appears as a broad band entirely surrounding tbe protoplasmic 
contents of tbe cell. Tbat tbis layer represents merely tbe swollen 
inner wall of tbe motber cell seems doubtful, for it appears very 
intimately related to tbe enclosed protoplasm, wbicb by tbis time 
bas sbrunk away from tbe motber wall. It seems ratber to be 
an entirely new wall developed in anticipation of tbe formation 
of microspores. Tbe cell plates now split, and walls are laid down 
in tbe nsnal manner (fig. 37). Tbese walls tben apparently swell, 
become continnons witb tbe tbick enveloping wall (fig. 38) and 
assume tbe same reaction toward stains. In tbis manner tbe cells 
of tbe tetrad become separated from one anotber and from tbe 
ontside by a tbick, transparent wall, wbile tbe original mother cell 
membrane, already very thin and distorted, gradually disorganizes. 
Tbe cells of tbe tetrad, wbile still enclosed by tbis tbickened layer, 
begin to form tbe walls wbicb are present in tbe mature micro¬ 
spore. Tbese are sometimes evident in places wbere an enclosed 
cell bas become separated from tbe tbickened wall. Tbe pollen 
grains are eventually liberated by tbe dissolution or breaking down 
of the enveloping wall. 
Tbere can be little doubt tbat tbe peculiar structure described 
here, and found in Pinus, is in tbe nature of a cell wall and is 
not merely aviscid or liquid substance, and an examination ofliving 
material of J. communis fully confirms tbe observations made on 
fixed and stained preparations. Yet wbile in Pinus it “is left 
bebind as a definitely outlined wall after tbe escape of tbe spores” 
and “the empty mother-cell witb its four chambers is often met 
witb”, in Juniperus it is not of such a permanent character 
