430 THE JOURNAL OF BOTANY 



authentic examples so named by him. Dumortier (Fl. Belgica, 

 27, 1827) seems to have first placed De Candolle's plant under the 

 genus Limonium ; we should therefore label specimens L. bellidi- 



foliam Dum. 



The European distribution of our plant, as given by Nyman 

 (Conspec. FL Europ. 1878-82, and Supp. 1890), includes Spain 

 (Catalonia), Balearic Isles, South France, Italy, Sardinia, Austria- 

 Hungary (Dalmatia), Turkey (Macedonia), Rumelia (Thracia), 

 Rumania (Dobrudscha) , Greece and South Russia; lie separates 

 our British plant as an endemic subspecies, " S. reticulata Sm.," 

 found only in Norfolk and Lincoln. 



To find a reason why Nyman should so label our form, I ex- 

 amined numerous examples of the species from all the countries 

 mentioned above (with the exception of Rumelia and the Balearic 

 Isles), and the results were as follows. In Spain, France, Italy, 

 Sardinia, Austria-Hungary, Turkey, and Rumania the plant is 

 usually identical with our native species, varying greatly, as in 

 England, as regards barren branches, laxity of spike, &c. 



In Greece and Russia the plant very often assumes a different 

 aspect, and bears short dense spikes of smaller flowers with more 

 rounded calyx-lobes (indeed Halacsy, Conspec. Fl. Graec. iii. 22, 

 1904, describes them as ovate), and is thus (particularly in Greece) 

 nearer to the S. Iconia of Boissier & Heldreich. The latter, 

 however, a rare plant of Asia Minor, has a different habit of 

 growth, a distinctive hairy and small calyx, and other points of 

 divergence. In Greece and Russia one finds that a very great 

 variety of forms exist, which provide links between the extreme 

 state mentioned and our English plant, and I am led to the 

 conclusion that British examples, whilst often possessing larger 

 flowers than those from the Continent (possibly caused by a 

 difference in latitude), cannot be held to be a separate subspecies, 

 as noted by Nyman, but are merely states or forms of this vari- 



able species. 



bellidifi 



not be out of place here : — Wells, England — very few or no sterile 

 branches, spikes short dense; Hunstanton, England — numerous 

 sterile branches with similar spikes ; Spain — sterile branches 

 many, spikes lax; South France — sterile branches many or few, 

 spikes short dense or longer and laxer; Italy — sterile branches 

 many, spikes more or less dense, scale often leafy ; Sardinia 

 and Austria- Hungary — sterile branches many, spikes rather lax ; 

 Turkey and Bumania — sterile branches many, spikes lax, calyx- 

 lobes more rounded; Greece — sterile branches many, spikes very 

 short dense, calyx small with lobes rounded ; South Russia — very 

 variable, many sterile branches, lax spikes, calyx of normal size 

 with lobes triangular-acute, or sterile branches few or none, spikes 

 dense, calyx small with rounded lobes. 



A variety, patens, of which I have examined a type-specimen 

 in Herb. Boissier is mentioned by him (DC. Prodr. xii. 661) as 

 occurring in the Altai region near Loktiew, Siberia ; I have not 

 seen examples from other countries. 



