Nichols, A morphological study of Juniperus communis var. depressa. 211 
but disintegrates rapidly after the release of tbe pollen grains, 
conveying tbe impression tbat it is composed of some gelatinous 
substance. Incidentally it may be noted that similar tbickened 
walls have been obseryed by tbe writer in the living pollen motber 
cells of Picea excelsa, and, as Miss Ferguson intimates, it is 
probable tbat tbese structures are of general occurrence tbrougbout 
tbe higher plants. 
Düring the time which intervenes between tbe formation of 
tetrads and pollination tbe spores round out and increase in size 
at tbe expense of protoplasmic material derived mainly from tbe 
disorganizing tapetum and inner layer of tbe sporangium wall. 
Tbe mature pollen grain (fig. 39) is approximately spberical, about 
20 /x in. diameter, and possesses a tbin, transparent intine and 
a sligbtly tbickened, pigmented exine. Tbe cell contains a single 
centrally situated nuclens, and scattered tbrougb tbe cytoplasm, 
often nearly obscnring tbe nucleus, are an abundance of large, 
usually compound starcb grains. 
In regard to tbe presence of protballial cells, such as occur 
in man} 7 Gymnosperms, it ma} T be stated with almost positive cer- 
tainty tbat tbese structures are never formed in tbe Cupresseae. 
Tbat tbis is true of Juniperus was first pointed out by Stras¬ 
burg er (1892). Tbe material studied by tbe writer sbows two 
peculiar features wbicb at first sigbt suggest tbe presence of a 
protballial cell. It is observed tbat sbortly after tbeir formation 
many of tbe spores contain two nuclei. A tborougb search, bow- 
ever, fails to reveal any indication of mitotic figures, such as would 
be expected bad both nuclei arisen in tbe pollen grain by tbe di- 
vision of tbe primary nucleus. Moreover tbe two are apparently 
alike in every respect, wbile tbe spores tbemselves are invariably 
somewbat larger tban tbe normal and are frequently constricted 
at tbe middle. In view of tbese facts it seems probable tbat tbe 
binucleate condition is brought about by tbe failure of a cell wall 
to develop between two nuclei of a tetrad, and tbat a consequent 
twinning of two microspores results. In tbe same sporangia witb 
tbese abnormal pollen grains are found otbers in wbicb there 
appears wbat migbt easily be mistaken for a disorganizing pro¬ 
tballial cell, — a dark, lenticular structure, seemingly closely 
appressed to the wall of tbe cell. But there is little doubt tbat 
tbis is merely an artificial condition indnced by a slight invagination 
of the spore wall. 
Tbe primary nucleus of Juniperus undergoes no division until 
after pollination. In tbis respect Taxus, Cupressus and Juniperus 
differ from all other Conifers, so far as is known. 
Tbe wall of tbe mature microsporangium is composed of a 
single layer of prismatic cells wbicb are elongated in a direction 
parallel to tbe slit by wbicb tbe anther debisces. A tbin layer 
of cytoplasm lines tbe cell walls, and tbe nucleus in many cases 
bas not yet disorganized. Tbe cell walls are comparatively tbin, 
but on tbe lateral walls are rib-like bands of thickening (figs. 13, 14) 
wbicb stain bluisb with cyanin, wbile tbe wall itself is erythrophilous. 
, 14* 
