160 Blackburn and Harrison. — The Status of the British 
2. Recognized plant hybrids generate ill-developed pollen ; is that of 
the Rosae equally imperfect ? 
Although the cytological work was begun the earlier, the second 
research has been published first (Harrison 1920 ) ; still, we must point out 
that a risumi of the cytological results was given in a paper read by one 
of us at the British Association meeting in August, 1920. Of the latter 
paper the present account is an expanded statement. 
Straightforward as the questions propounded seemed to be at first 
sight, difficulties were soon encountered, for apomixis occurs in its facultative 
guise in practically every rose species or microgene. Thus, if the bulk of 
our roses are hybrids, as the results of both researches indicate, in the Rosae 
as in other genera, Taraxacum , Erigeron , Hieracium and the like, hybridfty, 
polymorphism, and apomixis are related to each other in the way of cause 
and effect. The elaboration of this point will be reserved for the discussion 
at the end of the paper. 
Material , Methods , &c. j 
All of the material examined, with the exception of that of Rosa | 
arvensis and part of that of R. rubiginosa , was collected in Northumberland ' 
and Durham and was determined by ourselves. Our determinations we 
can guarantee to be accurate because the necessary examinations were 
carried out on the growing bush — a sine qua non when dealing with critical 
microgenes. 
In our earlier work the young buds were fixed in Flemming’s fluid, 
but later other fixatives were tested, to be abandoned finally in favour of 
Carnoy’s fluid, which, for the end in view, proved the best reagent. 
The sections were cut to a thickness of 5 n, stained with Heidenhain’s 
iron-alum-haematoxylin, and mounted in balsam without a counterstain. 
Very early indeed we became aware that in the genus microspore [ 
formation occurred of two types, one perfectly normal and the other 
exhibiting irregularities, varying with the microgene under investigation, 
but nevertheless always of the same order. Furthermore, it appeared that 
the abnormalities occurred in forms with high chromosome numbers, but 
were almost absent in those with lower complements, i. e. in microgenes with 
a haploid number of 7 or 14. Incidentally, it will be evident that Strasburger 
was in error in fixing the fundamental chromosome number in Rosa as 8, 
the correct figure being 7. 
The most frequent form examined with this latter haploid number was 
one of the Systylae, Rosa arvensis, and upon that species, aided by 
reference to R. pimpinellifolia , we shall found the description of events in 
the meiotic phase of a normal rose to be utilized subsequently in comparisons 
with anomalous forms. 
