37 
has the lax panicle and elongate branches common to the two vars., 
but like the first it bears only a single tubercle, which is of small 
size and gradually tapering into the midrib of the petal, as is some- 
times the case in var. subcordatus , but the fruit petals are small, 
and ovate-deltoid or triangular-ovate, as in var. elongatus, of which 
an account by Dr. Trimen will be found in the “ Journal of Botany,” 
1873, p. 237. 
R. crispus, var. trigranulatus , mihi. “ Swanbister, Orphir, Orkney, 
1874 ; Brodick Arran, 1873; Seafield, near Kirkcaldy, Fife, |1 874. I pro- 
pose the name of trigranulatus /or a littoral variety of crispus, which has 
a very dense panicle, with short, adpressed branches and small triangu- 
lar-ovate fruit-petals, each of the three bearing a conspicuous tubercle. 
This form is reproduced from seed. The panicle has somewhat the 
appearance of R. domesticus, but in no other point does the plant ap- 
proach that species. It agrees with var. elongatus in the shape of the 
petals and in cell, three bearing tubercles, but it differs in the compact 
panicle and the radical leaves being very strongly crisped. The 
Honourable J. L. Warren in a letter suggests that possibly the position 
of elongatus at Putney growing in the tidal mud and under water at 
every tide may cause the root-leaves to uncurl and flatten out, and 
he thinks that specimens which Mr. H. C. Watson got with him at 
Putney, and is now growing in his garden, are reverting to rather 
crimped root-leaves. Mr. Watson supplied me with ripe seeds of 
elongatus , but none of them have germinated. I have frequently found 
the seeds of docks lie dormant for two or three years, if they be not 
sown immediately on ripening, which is best.” — J. T. Boswell. 
Rumex Hgdrolapathum , Huds. “ Mr. H. C. Watson has sent me 
a number of examples of the radical leaves of R. Hgdrolapathum from 
the Thames side, near Moulsey, and Surrey Canal, from Woking to 
Byfleet, collected in 1874. These leaves show the transition from 
the attenuate to the abrupt and even subcordate form of base. Some 
of the specimens, particularly one from a plant by the Thames side 
just above Moulsey Hurst, has broadly oval-ovate leaves, with sub- 
cordate bases assigned to R. mqximus, Schreb. Unfortunately this root 
was so placed as to be on a sort of stand for fishers, while the Thames 
was low in summer ; thus no fruit-stem was perfected, and nowhere did 
Mr. Watson find fruit-perianth so much dentate as in Sussex maximus. 
For my own part I lay more stress on the shape of the enlarged petals 
than on their dentition, and I have not seen any British specimens 
except Mr. Warren’s Sussex ones, which approach the Continental 
maximus in shape and size. I have not yet been able to obtain British 
specimens of R. maximus , and should be much indebted to any bota- 
