346 Olive. — Sexual Cell Fusions and Vegetative 
nuclear sap or else anastomoses with other strands. Although a consider- 
able amount of nuclear sap seems to have been already formed in this 
instance, the nuclear membrane cannot yet be distinguished. As in the 
cases described above for the gametophytic hyphae, the chromatin strands 
which radiate from each centre are plainly constant in number, being about 
eight in each of the three nuclei shown. The fourth and missing nucleus in 
this cell lies immediately below the one shown in the tip of the hypha. 
The large size of the nuclei of Triphragmium is emphasized by a com- 
parison of Fig. 19 with Figs. 20 and 21, which represent the close of the 
conjugate divisions in the aecidium of Uromyces Scirpi. All three of these 
figures are drawn to the same scale. The latter are too small to show the 
details of the telophases ; but the centres, chromatin strands, and central 
spindles can be made out in properly-stained preparations, and these features 
will be seen to correspond in every way with those described for the larger 
and more favourable nuclei. 
It will be remembered that Richards (’ 96 ) found at the base of the 
aecidium-cup of the Rust on Ranunculus septentrionalis a single large 
c fertile hypha ’ which contained many nuclei and which budded out to form 
finally the basal cells. Massee (’88) had previously figured a large several- 
nucleated cell which he termed the ‘ oogonium ’, at the base of the aecidium- 
cup on Ranunculus Ficaria. Blackman and Fraser (’OS) also show drawings 
of Puccinia Poarum (Figs. 15 and 16) in which three and four nuclei are seen 
in division ; but they regard such cases as abnormal. 
In the species occurring on Cicuta ( Uromyces Sciipi ), as well as in six 
or eight other species, I have found multinucleated cells, sometimes only 
one, or again many, generally scattered more or less irregularly at the base 
of the aecidium-cup. Fig. 21 illustrates one of these cells, which contains 
four nuclei in an advanced state of division. The parallelism of the four 
spindles in this case is striking, only one of the nuclei being placed at 
a slight angle to the others. Notwithstanding this slight displacement, the 
two pairs of nuclear figures may be still regarded as maintaining their con- 
jugate arrangement. In the later development of the fungus, when two 
nuclei only occupy the cell, the close proximity and parallelism of the two 
figures appears to be maintained with the greatest regularity. 
Another four-nucleated cell is shown in the next figure (22), from the 
base of the aecidium on the thistle. Nuclear division has apparently just 
been completed, since the nuclei show obvious signs of undergoing recon- 
struction. Three of the nuclei, for example, still have the chromatin 
massed about their centres. 
Much more striking cases of multinucleated cells occur in Figs. 23, 24, 
and 25, two of which (23 and 25) are from the base of the aecidium on the 
thistle. Besides the nine poorly differentiated nuclei shown in Fig. 23, 
three possessing the same characteristics and undoubtedly belonging to the 
