234 riTTONiA. 



and have so failed in knowledge of the habit and sensible 

 properties of the subjects of their, study; and habit ^vill 

 sooner or later be received as of more siiruificauce in delimit- 



ing genera, than technical character. In these very natural 

 orders, habit must largely determine the limits of genera. 

 Bat it is in dealing with these families, that our most dis- 

 tinguished systematists of the herbarium, have been driven 

 to tiie magnifying of every little difference in the f<n-m of 

 corolla and posture of stamen, into a characterj in order that, 

 with their imperfect resources, they might have definable • 

 genera. For example: the Pacific American species of Trich- 

 osfcma with the Atlantic Isanthus, form a group so eminently 

 natural, that a blind man would generalize them into one; for 

 they all have one and the same peculiar odor, which is not 

 like that of any other plants of this or any other family; and 

 they are all so much alike in habit, inflorescence, and every- 

 thing else save the shape of the corolla, that, in no natural 

 system of classification can they be resolved into two genera. 

 Even LinuDGUS, supposed to have been the most artificial of 

 systematists, did not force these plants asunder. And in 

 Trichosfema, with Isanthus included, there is no wider 

 range between irregularity and regularity in the form of 

 the corolla,^ than is displayed in Pentstemon, or even in 

 Caprifolium, 



No botanist, whether of much or little experience, coming 

 into California from Europe or eastern America, and meeting 

 With our native kinds of sage, would entertain a doubt of their 

 being perfect Salvias, botanically speaking. But they have 

 a calyx differing somewhat from that of the best types of 

 Salvia, and only one of the connectives of the anthers bears 

 a pollen-cell, the other connective being reduced or obsolete. 

 There is, however, no dearth of admitted S-ilvias in which one 

 of the connectives is reduced and sterile. On some such 



^ The figure of L ccenilein^ in Michaux is very misleading and errone- 

 ous, showing. as it does, a corolla with broaJ rouuded lubes like those of 

 nMyosotis; but th:U of Dlllenlus (Elth. t. 3S5), thou^'h on the whole 

 mucli lei5 faulty, parhaps shghtly exaggerates the irregularity. 



