SYSTEMATIC BOTANY. 



learn, siuce its characteristics are Avell marked. "We will suppose that 

 you have before you a very common hlue-flowered shrubby plant belong- 

 ing to that order, a single flower of which is shown in Fig. 58. Knowing 

 the order, you turn at once to p. 38, and begin to use the key to the 

 genera. 



56. Carefully removing all the petals, the stamens and pistil appear as in a. Fig. 59. 

 Tiie filaments are united for the greater part of their lengtli into a tuV)e which incloses 

 the ovary as a sheath docs a knife. Of the three sections in the key, then, it is evident 

 that the second is to be taken. Since some of the anthers have shed their pollen, and 

 others have not, it is safe to say they are of two kinds — a bud will show the difference 

 better (b, Fig. 59. ) Moreover the leaves are digitate, and have more than three leaflets. 

 "We therefore conclude that tiie third genus is the one. Our plant is slightly shrubliy, 

 so we pass oveir the first heading in the synopsis of species. Of the second and third 

 headings the last seems the most likely to lead us aright. Our flower is blue, so we have 

 to choose between the second and third species. The words " Slightly woody at the 

 base, " decide us in favor of Litpinus Douglasii, though we should examine more speci- 

 mens before being quite positive. 



57. Fig. GO represents a flower of a jjlant 

 common in the Redwood forests. Three or 

 more of the dull-colored flowers grow in an 

 umbel on a very short scape between a pair of 

 spreading radical leaves. Since the leaves have 

 parallel veins, and the parts of the flower are 

 in threes, Ave must use the Analytical Key 

 for Eudogens, p. 13. You will have no diffi- 

 culty in referring the plant to the Order Lilia- 

 CE^. To make the analysis of a plant in that 

 large order easy, the genera are grouped in the bursting anther ceUs 

 three Series. Reading the chai'acteristics of Series I, we find they do 

 not correspond with those of our plant, which has no floral bracts, which 

 has the stamens hypogynous instead of perigynous, the anthers extrorse 

 instead of introrse, etc. Comparing Series II with Series III, we decide 

 that our plant belongs in the former, since the perianth is not persistent, 

 and the flowers are not in racemes or panicles. § 1, in Series II is wrong, 

 for our i^laut has no leafy stem. Since the perianth segments of our 

 flower are dissimilar we try § 3, under which we refer our plant to the 



60- riower of ScoHopuB. a, ono 

 of the stamens nifiRuitied, showing 



