142 ORIGINAL ARTICLES. 



well as nearly complete sets of our then cultivated varieties, pure or 

 hybrid, from four of our largest living collections, and thus acquired 

 a tolerable idea of the characteristic features assumed by hybrids in 

 this genus. Yet among the wild plants there was only one, in an old 

 collection of Roxburgh's, that had the slightest appearance of a hybrid; 

 and among European ones, the only instance I am aware of, is that 

 mentioned by Hewett Watson, of the Cornish hybrids, between JE. 

 ciliaris and E. Tetralix. So in the genus Diantlius, according to C. P. 

 Gaertner, artificial hybrids are very readily produced, and are more fer- 

 tile than those of almost any other plants, and yet wild hybrids are 

 very rare. Lecoq, it is true, speaks of hybrids between D. Monspessulasnus 

 and D. Seguieri as being very abundant in the Montdore, and certainly 

 these two species are, in that locality, very variable, but not more so than 

 I have observed them in the Pyrenees, Provence, &c, when growing 

 separately. 



The apparent' permanence given by cultivation to abnormal or in- 

 termediate races has afforded a plausible argument against the supposed 

 constancy of the limits assigned to species in nature. The manner in 

 which the Cape Pelargoniums, the South American Verhenas and Pe- 

 tunias, &c, have produced varieties without end, blending the original 

 species together in inextricable confusion, is well known ; and gardeners 

 reckon with tolerable certainty on reproducing, by seed, the numerous 

 varieties of our kitchen-garden annuals. But, as in the case of artificial 

 hybrids, these plants are then placed in an anomalous condition, in which 

 they are maintained by cultivation only. Restore them to the condi- 

 tions of a wild growth, leave them exposed to all those obstacles which 

 nature opposes to their multiplication, and they will soon yield to the 

 more hardy or more favoured genuine forms, and gradually perish with- 

 out being reproduced. This temporary character, when wild, may be 

 observed in all the extraordinary aberrations from the common form, 

 however healthy the individuals may appear, such as Orchis pyrami- 

 dalis with spurless flowers, or Linaria vulgaris, with five spurs ; Helian- 

 themicm vulgare, or JVarcissus juncifolius, with linear or divided petals; 

 or Stellaria Holostea, with no petals at all, &c. ; they are none of them 

 perpetuated ; they cannot resist the immense chances there always are 

 against the offspring of any one individual plant ever coming to per- 

 fection.* 



To sum up the foregoing remarks : — When a plant is observed ap- 

 parently allied to some known species, but differing in one or more cha- 

 racters hitherto unobserved or unrecorded in that species, before deciding 



* As a familiar instance of the disproportionate chances against the success of any 

 individual seed in a wild pjant, take the foxglove {Digitalis purpurea). It will often 

 ripen 200 capsules, and even above twice that number have been counted on one 

 plant, and the number of good seeds I have found in one capsule have varied from 800 

 to 1200. Taking, however, the ave-age number of good seeds shed by every plant as 

 only 100,000 ; as the average number of foxgloves in a given district remains the same 

 year after year and century after century, we have only one plant coming to perfection 



