142 
THE GARDENING WORLD 
February 18. 1905. 
their works are scattered through many volumes and separate 
publications, so that unless the genus had been critically dealt 
with the student who wished to follow the nomenclature of 
the " London Catalogue ” would have been more or less com¬ 
pletely puzzled. 
We are pleased to note that R. hibernica and R. involuta are 
considered as separate and distinct entities, instead of as 
formerly being regarded as mere varieties of R. spinosissima. 
Concerning tins matter the author admitted that like other 
naturalists he was unable to define what a species was. At 
the same time he thought that many plants could easily be 
distinguished in a living state though it was difficult to dis¬ 
tinguish them by means of dried specimens alone. It was 
also his firm belief to decide upon the probable distinctness 
of plants before attempting to define them, and to make the 
plants themselves afford the distinguishing characters rather 
than that the characters should define the species. Such being 
the case it is well to allow those forms of plant life that are 
relatively distinct to be named by a generic and specific name 
rather than regard them as varieties and sub-varieties in a 
complicated way. 
Referring to R. hibernica, the editors say it is generally con¬ 
sidered a hybrid between R. spinosissima and R. canina, 
although they have not made bold enough to classify the forms 
in accordance with the idea. The variety R. h. Grovesii, both 
from its form and its isolation from other forms of R. hiber¬ 
nica. point pretty conclusively to the fact that its presence in 
such an isolated position with the parents is pretty good 
evidence that it is a hybrid. The editors also include a large 
number of forms as varieties under R. involuta, which are also 
regarded by authorities on the Rose to be hybrids. The 
number of forms placed under this species has been brought 
greatly up to date, the idea of the editors being to give a re¬ 
corded description of the various forms now listed in the 
‘‘London Catalogue.” 
We note also that R. villosa has been retained as the name of 
a British Rose, but although Linnaeus was the author, we fear 
that it included R. tomentosa as well. There are certainly 
two species indicated by these names and readily distinguished 
by students in the field, and while R. tomentosa is fairly well 
scattered over the length and breadth of the British Isles, the 
other is distinctly confined to the north, and has its head¬ 
quarters in the Highland glens, where it reaches its greatest 
abundance and perfection. 
The group of Dog Roses has now been broken up into three 
species instead of being included under one. In the old re¬ 
visions of the Dog Rose there were more than thirty varieties 
described as natives of Britain. We think it a great improve¬ 
ment in the separation of a number of allied forms under R. 
obtusifolia, as the species had been described by a. Continental 
author many years ago, but more or less ignored by authors 
in this country. Another curiously mixed and complicated 
group has been separated under the specific name of R. glauca, 
the various forms being described as varieties of it. In our 
opinion the whole of this group, like those described under R. 
hibernica and R. involuta, consist of hybrids. Each of the 
forms is briefly and succinctly outlined, but the botanist in the 
field will find large numbers of specimens which do not strictly 
agree with the descriptions and yet may be classed under the 
varietal name as sub-varieties. The combinations of forms in 
each of the so-called varieties, together with the environment, 
are strongly suggestive that they are of hybrid origin. 
R. canina, as it now stands in this work, still includes seven¬ 
teen varieties, each of which has a distinctive name. The 
student who is only making an attempt to get acquainted with 
the Roses may have difficulty in finding out which is the 
typical form or type of the species, but as he proceeds he will 
find that all the forms he discovers will arrange themselves 
under one or other of the named varieties, so that there is 
really no typical form of R. canina. The latter is merely in¬ 
tended to comprehend the whole group, and if the student is 
satisfied to place the Roses he may find under one or other 
of the specific names that will be simple enough. 
What applies to R. canina also holds good with regard to 
every other British Rose, of which thirteen species are here 
recorded with nearly all of the known British varieties classed 
under their respective headings. At least three of these 
specific names in our opinion are mere names that serve to 
string together a number of forms which are natural hybrids. 
We have no fault to find with tins means of classifying them, 
however, as it is adopted by various other first-class authorities 
in this country. We consider that it simply serves the purpose 
of classifying and simplifying the great number of forms to 
be met with. 
Excellent as was the author's description and classification 
of the Hawkweeds (Hieracium), a- large number of forms have 
since been found in the British Isles, and accordingly the 
editors thought it was desirable to substitute a fresh account 
in order to give an intelligible view of the British Hawkweeds 
from a modern point of vi'ew. Mr. F. J. Hanbury, who re¬ 
cently published the “ Flora of Kent,” Iras been very active 
in tins particular genus, so that a fresh account of the genus 
has been drawn up under his supervision. The British species 
of Hieracium now stand at ninety-seven, exclusive 1 of the very 
numerous varieties. 
For many years past Professor Babington was the recognised 
authority on the British species of Rubus or Bramble, and even 
the seventh edition of his work contained an excellent account 
of them. The editors could scarcely permit themselves to in¬ 
terfere with the master’s handling of this complicated genus, 
and they retained the genus as it stood. 
In order to bring the account alongside of modem know¬ 
ledge, the permission of the Rev. W. Moyle Rogers was 
obtained for a reprint, giving a conspectus of the genus as it 
is given in his “ Handbok of the British Rubi.” This has 
accordingly been placed at the end of this book as an appendix. 
Babington’s “ Manual ’’-will therefore hold its own in being up 
to date, even in the matter of Brambles. We can therefore 
fully commend this new edition to the attention of students 
and those who desire to study the British flora, even if they 
have any previous edition, as this is" thoroughly up-to-date. 
Scilla peruviana. 
(See Supplement.) 
The typical form of this plant is possibly lilac or pale blue, 
but even in a state of nature it is a variable plant, and occurs 
also with reddish and whitish flower's, but the finer forms now 
to be' found in gardens have deep blue, violet or violet-purple 
flowers. The clump of plants we illustrate is that in the col¬ 
lection at Ivew, where two varieties were noticeable, but most 
of them had flowers which were violet-purple in bud, and varied 
with a darker or lighter shade of that colour when fully ex¬ 
panded. The Scillas have flowers of different forms according 
to the species, but those of S. peruviana have their segments 
spread widely, giving them a starry appearance. The stamens 
have deep blue filaments and yellow anthers, the ovary being 
brilliant blue. When these flowers appear above ground, the 
raceme is in the form of a corymb, but later on becomes shortly 
pyramidal and then racemose. The flowers range from fifty to 
110 , and are borne on stout, scapes 12 in. to 18 in. high when 
planted in the open ground. 
Several varieties have received special names, one of which 
is S. p. alba, which has white flowers and was introduced in 
1 607, the same year as the type. Another distinct one is S. p. 
discolor, having yellow-brown flowers and blue filaments and 
ovary. This was introduced from Portugal in 1843. S. p. 
glabra has deep lilac flowers, and the leaves are not ciliate at 
the edges, the varietal name alluding to this. Still another 
variety was introduced from Sicily in 1873—S. Hughii, a more 
robust form than the type, with leaves 11 in. to 2 in. broad. 
The pedicels, bracts and the scape itself are tinted with red. 
All of these varieties bloom about the end of May and the 
beginning of .Tune. 
Readers will be puzzled with the name S. peruviana, which 
would indicate that it comes from Peru inasmuch as the name 
