30 Mr Bur kill, Trifolium pratense var. parviflorum. 
to the heads and short pedicels to the flowers. This last is 
Trifolium pratense var. parviflorum , and has the following 
synonymy : — 
T. pratense var. parviflorum, Babingt., Manual Brit. Bot., 
ed. 1 (1843), p. 72 ; Lange in Oeder’s Flora Danica, t. 2782. 
T. brachy stylos, Knaf in Lotos, 1854, p. 237. 
T. pratense var. pedicellatum, Knaf ex Celakovsky, Prod, 
d. Flora von Boehmen, iii. (1875), p. 669. 
T. pratense , forma T. brachy anthe mum (3 heterophyllum, Rouy 
in Rouy et Foucaud, Flore de France, v. (1899), p. 120 (published 
as Ann. Soc. Sc. Nat. Charente-infer.). 
Babington’s type-specimens from Elgin, as well as others from 
Plymouth and Walton-on-Naze, and a type of Lange’s figure 
have been accessible to me in the Herbarium at Cambridge ; a 
type of Knaf’s name, collected by Auerswald in Bohemia, has been 
seen in the Botanical Department of the British Museum of Natural 
History, South Kensington ; at the Royal Gardens, Kew, are speci- 
mens collected at Fairmile in Surrey, at St Leonards, at Tonbridge 
Wells, and at Elgin, from the herbaria of Borrer and H. C. Watson, 
and from near Bordeaux, collected by C. des Moulins ; and I have 
myself collected it at Hunstanton in Norfolk, Gatton Park in 
Surrey, Waltham St Lawrence in Berkshire and (in company with 
Mr G. Nicholson) near Heiligenblut in Carinthia — on each occasion 
a single root. All these specimens agree very closely. 
The first definition of the variety parviflorum runs : “ heads 
more or less stalked : calyx-teeth as long as, or longer than, the 
corolla,,” and is correct as far as it goes. Celakovsky’s description 
is “Ahren grosstentheils gestielt ; Bltithen langer oder kiirzer 
gestielt ; Deckblatter theilweise ausgebildet ; Griffel ktirzer als 
die Staubgefasse.” But the following is fuller and more in accord 
with the specimens : — plant not robust ; heads more or less stalked ; 
bracts sometimes developed ; corolla in the mature flower crumpled 
at the base within the calyx and not exceeding the longest of 
the calyx-teeth ; pistil becoming foliaceous, the ovarial part linear- 
lanceolate, and often open above ; ovules more or less aborted. 
Examination of buds not ready to expand reveals no crumpling 
of the corolla ; so that this evidently takes place in the rapid 
growth of the tube which precedes the expansion of the flower; 
and it is impossible to resist the assumption that the unusual 
size of the ovary and the narrowness of the mouth of the calyx 
are the causes of it. 
Phyllody of the ovary to a greater degree than in typical 
parviflorum is not uncommon in Trifolium pratense ; less modifi- 
cation in this direction I have found in a plant from Glen 
Clova, Forfarshire, where the peduncle and pedicels were un- 
developed, but the corolla crumpled and the ovary elongated, 
