Rose Forms as determined by their Cytological Behaviour. 1 6 r 
II. The Meiotic Phase in the Species behaving normally. 
[a) Rosa arvensis , Huds. 
As this is not intended to be a critical survey of the cytology of Rosa } 
but is rather an inquiry into the status of its British members, the series 
of sections figured here simply displays representative stages in pollen- 
formation. sufficiently complete in themselves to show that in the truly 
sexual roses the general course of events differ not at all from that followed 
in typical phanerogams. 
The stages chosen for figuring commence with PI. IX, Fig. i, in which 
we depict the resting nucleus just preparing for synapsis by an increase in 
the bulk of its chromatin. Onward from this there is a progressive massing 
of the chromatin to one side of the nucleus, where it is arranged, more or less 
neatly, between the nucleolus and the wall. When synapsis is at its maximum 
intensity, the chromatic clump becomes exceedingly dense and granular, as 
is shown in PL IX, Fig. 2. As the synaptic knot unravels, obviously looped 
sections are thrown out into the nucleus (PI. IX, Fig. 3). Next the loops 
extend until the whole nuclear cavity is filled, and the ordinary hollow 
spireme stage is reached. The beaded spireme then thickens, and, as it does 
so, reveals itself as made up of seven long loops. The succeeding stage 
shows the units of the bivalents twisted round each other (PI. IX, Fig. 7). 
Henceforth the chromosomes concentrate and thicken with an accompanying 
increase of their staining powers and a diminution in the case of the nucleolus 
to more or less cytoplasmic values (PL IX, Fig. 10). Subsequent to this an 
irregular multipolar spindle (PL IX, Fig. 11) of a transient nature arises, 
passing by imperceptible steps into the typical bipolar form. At this 
point the somewhat oval chromosomes reach the spindle (PL IX Figs. 12 and 
13) when, as is quite plain, they lie with perfect regularity on the equatorial 
plate ; now, too, the nucleolus has disappeared. The separation of the 
bivalents takes place with the utmost regularity, and the chromosomes, 
without exception, reach the poles simultaneously (PL IX, Fig. 14). 
No hint is given here of hybridity ; the perfect assembling of the 
chromosomes on the equatorial plate, their synchronizing in doing so, the 
neat arrangement of the separated chromosomes on the spindle, as they 
pass to the poles, force upon one’s mind the certainty that, in this species 
at least, no hybridity exist either patent or latent. There is absolutely no 
suggestion of lack of pairing, irregularity in passing to and from the 
equatorial plate, so obvious in the heterotype division of Drosera obovatU , or of 
the first maturation division of our Lycia-Nyssia and Oporabia moth hybrids. 
In PI. IX, Fig. 15 the interkinesis between the heterotype and the 
homotype divisions may be seen with the chromosomes represented so that 
no longer can the chromosome number be made out. At this stage the 
nucleolus comes once more into view. 
M 
