VOL. XVIII. (3) NOTES ON HELOSCIADIUM 233 



Since this paper was published, I have examined many 

 hundreds of herbarium specimens, and, what is more im- 

 portant, studied the species frequently in the field. Certain 

 conclusions have slowly formed in my mind as a result of 

 this study. The first is that the varieties already mentioned 

 represent only a few out of numerous forms, all equally capable 

 of description, which the species assumes. The varieties 

 named happen to have been pitched on by collectors in past 

 years, and described : and so they have got into our hand- 

 books. They represent, in fact, certain stages, or steps, or 

 extremes, of variation in a few directions, but leave the student 

 uninformed of the large variety of directions and degrees in 

 which the species varies. The number of characters in which 

 variation occurs is, of course, great ; e.g., size of plant, vigour, 

 compactness, production of roots or rootlets at the nodes, 

 direction of stems or branches [i.e., prostrate, ascending, 

 etc.), shape and number of leaflets, length of peduncle, presence 

 and number of involucre-bracts, number of rays in umbel, 

 and so on. The degree of variation in each of these characters, 

 again, differs considerably. Now it is a fact that a large num- 

 ber of the possible combinations of these variations of character 

 may be found in a long series of specimens ; and if you were 

 to describe and name every distinct form, you would not 

 exhaust your opportunities for making and naming new- 

 varieties, and adding your name as author, under a very large 

 number. The forms of the plant, in fact, are " legion." 



A second conclusion is this : that you can often find 

 upon one plant, springing from one root, more than one of these 

 distinct forms. Here are some of the facts which bear out 

 this statement. Take such a character as that of " rooting 

 at the nodes." H. nodiflorum normally consists of a root from 

 which a more or less upright flowering stem proceeds ; and 

 there are numerous nearly prostrate side stems or branches 

 from the root, each of which may both flower and root. Such 

 a plant was one I gathered last year in a wet spot on shingly 

 ground at the foot of the cUffs at Beachley. The rooting 

 character may be present in all the side stems, of course, at 

 the nodes only ; even if for any reason the stems or branches 

 are not actually prostrate, if, e.g., they are carried up a hedge. 



