196 KEW PUBLICATIONS, 
of species lie steers a middle course "between extreme views. The pre- 
sent work is mainly made up of descriptions of an extremely similar 
character to those of the ^ Flore de France/ but those and the accom- 
panying critical remarks are of course written with the advantage of 
later knowledge. The present part is intended to be one of two. It 
goes down the scale, deviating veiy slightly from the ordinary classifi- 
catioii and the Candollean sequence of Orders, as far as LoranthacecBy 
the preface and all introductory explanations with regard to the tract 
of country included, being deferred to the concluding part. With re- 
gard to the limitation of species, except on the genus Rosa^ M. Grenier 
modifies here very slightly the views expressed in the larger work. 
The extracts we have marked have heen selected as bearing upon 
British botany, and as calculated to convey to those to whom his pre- 
vious writings are not familiar an idea of the scope and character of 
M. Grenier's observations, 
Draba verna, " I have not succeeded in testing by cultivation the 
value of the species formed by the division of Draba verna. For a long 
time I have sown seeds in the spring, and although germination took 
place, the plants were always destroyed in summer without coming to 
perfection, I have learnt later that the sowing should take place in 
autumn ; then the seeds germinate before winter, and produce flowers in 
the spring, so that the plants are, as M. Jordan has well observed, 
more truly biennial than annual," 
Drosera ohovata^ Koch. Apropos of the theory of the hybrid origin 
of this form, answering an objection of Fries, to the effect that no 
intermediates linking it with its supposed parents had been seen, M, 
Grenier writes, — " In the year 1850, I found in turf bogs of Portar- 
lier the answer to this objection, for I brought together a series of 
individuals passing by insensible stages from J), rotundifolia to longU 
folia, of which D. obovata constituted the mean term. Finally, I ob- 
tained the following series, abstraction being made of a crowd of inter- 
mediates, viz, 1. I), rotundifolia ; 2. _Z>. syper-rohindifolio-longifoUa ; 
3. D, loncff olio -rotundifolia ; 4. D, mper-longifolio-rotiindifQlla ; 5. 
jD. longifoUa. The D. obovata is then neither a variety of D. longi- 
folia, as Koch places it in the second edition of his ' Synopsis/ nor a 
true species, as he regards it in the first edition, but simply a hybrid, 
of which all the terms of the series passing from one of the types to the 
other are now perfectly known." 
