188 Rhodora [DECEMBER 
and these plants in the majority of cases show a reversion to the 
supposed parent types, which of itself, to our mind, is a positive 
proof of hybrid origin." No one will dissent from such a conclusion 
and it is therefore disappointing that the authors failed to tell us 
just which of the suspected hybrids gave these figures. They do 
report on 9 cases, the 3 above referred to in which seedlings “do not 
revert to the parent types" and 6 others in which they show varia- 
tion. But the thesis would be more convincing if reports had been 
included for the remaining 37 reputed hybrids. 
A serious doubt as to the finality of the conclusions in the paper 
must inevitably occur to those who have an intimate field-knowledge 
of the abundance in some of the upland districts of New Hampshire 
and Vermont of such thoroughly characteristic blackberries as R. 
frondisentis and R. abbrevians, shrubs with almost abnormally per- 
fect pollen for a Rubus, with seedlings true to type and both with 
finely developed fruit, for although the plate before us (Plate xxviii) 
shows woe-begone and discouraged little fruits on R. frondisentis, 
the large and abundant colonies in the swamps of northern New 
Hampshire bear splendid plump berries (as shown by manysheets 
of specimens indentified by Dr. Brainerd). If these are to be treated 
respectively as “ R. pergratus x setosus” and “ R. frondosus x setosus,” 
while R. elegantulus and R. vermontanus, of closely similar range and 
with amazingly imperfect pollen, are good species, why do not the 
hybrids occur generally throughout the coincident ranges of their 
supposed parents? R. pergratus is an abundant and much prized 
blackberry in many regions from Prince Edward Island to Cape Cod, 
Connecticut and Minnesota and R. setosus abounds in most regions 
from Nova Scotia and New Brunswick to western New England and 
the uplands of Pennsylvania. Yet in more than a quarter-century 
of intensive field-study and collecting of blackberries in New England 
and eastern Canada the reviewer (who has collected in a single season 
as many as 4000 sheets of Rubus and may perhaps be counted some- 
thing more than an “ordinary herbarium systematist," to quote 
Brainerd & Peitersen’s phrase) had never seen R. frondisentis until 
he turned his attention for two summers to the blackberries of the 
White Mountain region. Similarly he had never before met R. 
elegantulus, R. vermontanus and R. abbrevians. But all four are 
dominant and very distinct shrubs of the White Mountain region, 
although the reputed parents of the latter, R. frondosus and R. 
setosus, like the supposed parents of R. frondisentis, have much wider 
ranges. Brainerd & Peitersen assign R. frondosus to “Open fields 
and hillsides in southern New England. The form R. recurvans 
north into Maine, New Hampshire and Vermont," but 
they include i in R. frondosus not only R. recurvans but also R. phila- 
delphicus. The comprehensive species would thus have a range 
from Nova Scotia at least to western New England and Virginia, 
while reputed hybrids of it are cited from as far west as Illinois. The 
