1 74 Blackburn and Harrison — 77 ^ Status of the British 
is aborted pollen and a lack of intensity of the orange colour, when 
contrasted with that* of pollen of a superior grade, seen in the other 
canina and dumetorum segregates. At its very best R. hemitricha has only 
produced 1.5 per cent, of perfect pollen in its anthers. 
As in the Afzelianae, so in the Eucaniae, we regard the abnormalities in 
cytology as pointing directly to hybridity. 
(6') The Rubiginosae . 
Curiously enough, considered in conjunction with our matured con- 
clusions that nearly the whole of the roses are hybrids, the statement 
of Boitard in 1836 that many well-known rose growers had obtained seedlings j 
of Rosa ferox from R. rubigmosa seems very illuminating. However, our 
experiments showed emphatically that R. rubiginosa var. comosa , Rip., was 
facultatively and generally apomictical, so that seedlings from the same plant 
always agreed amongst themselves and with their parent in characteristics. 
This section is not well represented in this district, so that only two 
microgenes, R. comosa and R. apricorum , Rip., were studied, our local 
material of the forms being supplemented by some from Bedfordshire. 
In harmony with the Afzelianae and Eucaninae, the Rubiginosae proved 
members of the pentaploid series, having a somatic chromosome number of 
thirty-five reducing to twenty-eight for the heterotype division (see Text- 
fig. 3, d and j). That they mass themselves on the equatorial plate as in the 
other sections will easily be perceived from the same figure, and also from 
that on PI. X, Fig. 31, depicting a heterotype metaphase in profile. 
Of the forms previously under review the Rubiginosae most nearly 
approach R. hemitricha , so small is the proportion of pollen capable of 
germination developed. If the conditions in that plant be borne in mind, 
coupled with what prevailed in R. coriifolia , a good notion will be gained 
of the meiotic phase in the Rubiginosae ; thus unprofitable repetition will be 
avoided (see also PL X, Fig. 42). 
One occurrence must be singled out for special reference. In animal 
hybrids tripolar spindles in the maturation divisions are quite common, 
whilst in plants, on the contrary, although a tendency to multipolar spindles 
is the rule, they always yield to the bipolar type ; in R. apricorum tripolar 
spindles occur freely, and, what is more, persist, thus forming one of the 
factors in the wholesale failure of pollen in these microgenes (see PL X, 
Fig. 32). 
This, again, is one of the many features driving home and emphasizing 
the hybrid origin of practically every British rose. 
(d) The Tomeniosae. 
Since Rosa Sherardi , Dav., has been transferred to the Villosae, the 
section Tomentosae, as studied by us, is represented by two forms, R. 
sylvestris , Woods, and R. scabriuscula , Sm. 
