TWO BERKSHIRE VARIETIES. 349 



of Jordan's Diagnoses, using the word form in a different sense to 

 us, namely, to designate a state of the plant superior to a variety. 

 (See the preface to Flore de France.) 



Malva moschata L. var. Ramondiana Gren. et Godr. 

 As regards the above variety, I have to say that in my account 

 of M. moschata I followed Grenier and Godron in their Flore de 

 Fraiice, who give three varieties — laciniafa, with " tontes les feuilles 

 divisees en Lmieres etroites," the common plant in Berkshire, often 

 cultivated in the white-flowered form (I am told by a friend that 

 out of two thousand plants which he may have in cultivation in a 

 year, not one will show an undivided leaf) ; intermedia Gren. et 

 Godr., with "feuilles caulinaires divisees en lanieres etroites, les 

 radicales reniforme crenelees," rarer in Berkshire, but seen by 

 myself in several localities; Puimondiana Gren. et Godr., with 

 " feuilles toutes entieres dentees," which I have only seen near 

 Tilehurst. Some of the Tilehurst plants had all the leaves un- 

 divided (that is, typical Bamondiana), but the greater number had 

 the upper leaf or the upper leaves lobed, but not cut into segments; 

 in fact, the leaves, as we should expect, varied as leaves will vary. 

 But, as I have said, I have specimens from the locality which agree 

 with the description of Ramondiana, and M. Rouy, to whom I sent 

 one, accepts my determination. I also contend that it is better to 

 refer to Ramondiana plants which differ only slightly from it than 

 to put them to intermedia, from which they are much more widely 

 separated ; nor did I think it desirable to give a new name to 

 designate a varying series of forms. The above statement, to my 

 mind, meets the reviewer's objection, "That the plant I have called 

 Ramondiana does not agree with the description of the type." Mr. 

 Baker's remark in full was : " A very interesting variety of moschata. 

 This certainly must closely approach var. Fuimondiana, but I only 

 know this from the description. . . . The upper leaves are certainly 

 lobed." To this variety may also belong a plant referred to in 

 Woods' Tourist Flora, p. 62, in the following words: "Mr. Borrer 

 gathered in Somersetshire a variety of M. moschata which had the 

 stem-leaves as well as the radical leaves cordate-subrotund. It has 

 propagated itself in his garden for twenty years without alteration." 

 Woods' description of M. ynoschata would lead us to infer that the 

 plant he describes as the type is the var. ifitermedia, as the root- 

 leaves are said to be cordate-subrotund. It is interesting to find 

 that such "slight and variable characters" as those shown in the 

 cutting of the leaf may remain constant during many years of 

 cultivation. The names heterophylla and integri folia given to the 

 vars. of M, moschata by Lejeune and Courtois in the Cojnpendium 

 FlorcB BelgiccB, vol. iii. p. 14 (1836), and adopted by MM. Rouy and 

 Foucaud, are prior to the names intermedia and Ramondiajia of 

 Grenier and Godron, which only date from 1848. MM. Lejeune 

 and Courtois describe their integrifolia as having "folia omnia 

 rotundata inciso-lobata." MM. Rouy and Foucaud put var. Ra- 

 m,ondiana as synonymous with integrifolia; and their diagnosis 

 allows a much greater range of variation than I have assigned to 



