VOL, XVIII. (3) NOTES ON HELOSCIADIUM 237 



This raises the question : How distinguish H. repens 

 from H. nodiflorum var. longipedunculatum f. simulans ? 

 The crucial distinction hes in the fruit. In H. nodiflorum 

 the fruit is longer than broad (an occasional fruit is as broad as 

 long), very dark brown, almost black, with light coloured 

 prominent ridges ; its colour in section is dark. And this 

 holds true of all the forms which I place under H. nodiflorum, 

 including var. longipedunculatum and its f. simulans. But 

 in H. repens the fruit is smaller, broader than long, a light 

 pretty chestnut in colour all over, even on the ridges, and very 

 much the same in internal section. The exposed faces of the 

 fruit, mid-way between every two ridges, are puckered into 

 false ridges, so that there appear to be ten ridges to H. repens, 

 five of which are rather less marked than the rest. 



This is not, of course, the only distinctive mark of H. 

 repens. In it we have a good species ; but I do not think 

 that one permanent character, however strong, would justify 

 us in giving it specific rank. Besides the fruit character, 

 then, it has comparative stabihty of form. In a long series 

 of continental specimens I have found hardly any (if any) 

 which seriously depart from the strong pecuUar repens habit, 

 with its arching intemodes, even when growing in shade. 

 The same holds true of the plat in its occurrence at Skipwith. 

 The description of H. repens is : — Plant small, slender ; stem 

 prostrate, rooting at every node, arching. Leaves springing 

 erect or sub-erect from all the nodes, pinnate, total length 

 4-6, rarely 9 cm. Leaflets 9-1 1, suhrotund ovate in outline, 

 unequally dentate serrate, upper and middle leaflets sometime 

 lobed as far as the middle or slightly below, making a sub-bifid 

 leaflet. Peduncle of fully developed plant generally 3-4, 

 rarely 7 cm. lorig, 2-3 times longer than rays of umbel. Um- 

 bel-rays generally 5-6. Involucre of 3-7 bracts, persistent. 

 Inflorescence and leaves often approximately equal in length. 

 Fruit as above. 



It is (i) the leaflets (number and shape), (2) the constancy 

 of the involucre, as well as (3) constancy of habit, and (4) 

 fruit character, which, together, compel me to recognise 

 H. repens as a good species. Further research will probably 

 reveal other points. 



