1902] 



NEMO PHI LA 2 1 3 



Miss Eastwood has divided the small-flowered Nemophilas into eight 

 groups based upon the shapes of the corolla. These distinctions are too fine. 

 Such factors as age, the time of day, the degree of shade, or the amount of 

 nourishment received by the individual plant may easily make great differ- 

 ences in the appearance of pressed flowers. For example, in the morning a 

 plant may appear to have tubular flowers, while in the afternoon, when the 

 light is more intense, the same fl"owers may open wider and become tubular- 

 campanulate. If pressed in the morning condition, the flowers even when 

 soaked up will appear tubular, and correspondingly for those pressed in the 

 afternoon condition. As an instance of how dangerous such a close division 

 on this line may be, witness that Miss Eastwood has placed N. sepulta in 

 Section i, "Corolla tubular, minute," when that species almost invariably 

 has open-campanulate flowers. 



The position of the corolla-scales is also unreliable. In the living flowers 

 of most species the scales stand nearly perpendicular to the corolla ; so their 

 position in the pressed flower is accidental, or at most represents a very slight 

 inclination from the perpendicular in the living flower. One frequently .finds 

 in the same flower some scales folded towards the stamens and others folded 

 away from them ; so that in most cases very little significance can be attached 

 to their position.^ 



The shape of the scales is also exceedingly diverse, apparently much more 

 so than in the large-flowered species. Plants which agree in every other 

 particular often show quite different scales, and there is sometimes consider- 

 able variety on one plant or even in one flower. 



The writer has found no character which seems perfectly satisfactory for 

 discriminating species. In so variable a genus as Nemophila, and more 

 particularly in this most variable group, concomitance in variation should be 

 demanded before a form is named ; otherwise there is no limit to the number 

 of species we may make and our classification will become a burden. Some 

 of the species here reduced may prove to present such concomitance when 

 they are better known ; but at present the writer does not feel justified in 

 maintaining them. 



Unfortunately, the type specimen of iV. exilis is not a typical form of the 

 species as here described. The writer does not find the corollas of the type 

 "distinctly salverform," and he inclines to the opinion that their approxima- 

 tion to that form is due te the flowers shriveling before being pressed. 

 Flowers of even so open-flowered a species as N. Menziesii have been seen 

 where this had occurred with a similar result. If this character should prove 

 to have value, however, and if it can be connected with other characters, the 

 remaining forms should be separated as \^x\^Xy flaccida. Plate IV represents 

 the form published as N. nemorensis, which is the most common one about 

 San Francisco bay, and which is taken as typical of the group of forms which 

 make up the species. 



