38 BOTANICAL GAZETTE [JULY 
in each ovule, but generally only one and never more than two reach 
maturity. A little later the solid outgrowth of the exine becomes 
hollow, and the intine then grows through it, forming the tube, which — 
at once penetrates a short way into the nucellus, the single nucleus 
passing into the tube. Occasionally a tube is formed at. once on 
germination. 
The time during which the pollen tube remains uninucleate seems 
to vary a good deal, a tube with a single nucleus having been found _ 
associated with a four-nucleate prothallus (fig. 29), while in two or — 
three cases a tube containing three nuclei was found in an ovule — 
before the first division of the megaspore nucleus; one such case is _ 
shown in figs. 30 and 31. No differentiation of cytoplasm can be ~ 
seen here around the upper nucleus, doubtless the body cell nucleus, 
but the preparation is noteworthy as the only one in which a difference 
in size could clearly be seen between the two sterile nuclei, that nearest 
the apex of the tube (the “tube” nucleus) being considerably larger. 
Fig. 32 is-a detailed drawing of the tube, the position of which is 
shown in fig. 17. Its size is approximately the same as that of figs. 
30 and 31, but the structure is somewhat different. The two sterile 
nuclei are precisely alike in every respect, as they are in every prep- 
aration in which they occur, except in the one noted above and in 
one abnormal tube referred to later. The body cell is here sharply 
demarkated from the surrounding cytoplasm in which the other nuclei 
are imbedded. 
It should be mentioned here that the occurrence of only two nuclei 
in the tube at a stage somewhat later than this, as previously reported 
(10), has not been confirmed in any other preparations. It is curious 
that, in fact, the binucleate stage has never been certainly identified, 
as it is not improbable that the case just mentioned is really a three- 
nucleate tube, and that defective preparation is responsible for the 
appearance of only two. It may be added that the preparation there 
figured is from material collected very soon after the work was begun, 
and before the best fixing agents and oven periods had been satis- 
factorily ascertained. On the other hand, it may be that the uninu- 
cleate condition had persisted later than usual in this case, and that 
only two nuclei are actually present. In any event, it is certain that 
the two sterile nuclei must be cut off in somewhat rapid successio! 
