126 CALIFORNIA ACADEMY OF SCIENCES. 



tissue, and the characteristics of the filaments and their ap- 

 pendages, matters of acknowledged importance to the sys- 

 tematist, are sure to suffer obscuration, if not entire obliter- 

 ation, in the process of drying under pressure. Neverthe- 

 less, almost all which has been written hitherto upon Bro- 

 dkea and its allies has been written from the herbarium, and 

 all our authorities upon the group are foreign authorities. 

 No exception is to be made of botanical scholars belonging 

 to the Atlantic side of our own continent; for they are three 

 thousand miles distant from the habitat of these plants, and 

 as regards facilities for acquiring familiar and thorough 

 knowledge of tlie genera and species, possess little if any 

 advantage over authorities residing at London or St. Peters- 

 burg, Paris or Berlin. 



In the field there stand forth a few broad hints of generic 

 limitation which must, I think, impress every observer. We 

 have, for example, a group of perhaps a half dozen species 

 whose scapes are tall and weak and either actually twining 

 or else, by a marked tortuosity, expressing a demand for 

 extraneous support. And there is another group, more 

 numerous in species, whose scapes are short and rigidly 

 erect. But the external dissimilarity does not end here. 

 The voluble or tortuous kinds bear compact umbels of small 

 flowers; the stiff-stalked species have loose umbels of large 

 flowers; and, moreover, the two groups, as we for the pres- 

 ent call them, have each its own pattern of a perianth; some- 

 thing in the outline of that organ which, though nearly im- 

 possible to define in Avords, is recognized at a glance by the 

 botanist's eye, if he have the fresh flower before him. Now 

 if the individual perianths of the two kinds be carefully ex- 

 amined, other differences easily definable reveal themselves. 

 The weak-stalked, small-flowered assemblage of species 

 have uniformly a thin, somewhat inflated perianth-tube with 

 the body of which the filaments are so perfectly coalescent 

 as to disappear from the wall of the tube altogether. In the 

 stifl-stalked, loose-umbeled group the perianth-tube is of 



