1886.] BOTANICAL GAZETTE. 199 
Many species of Eriogonum would searcely be showing above 
ground by the end of May. In June, 1884, after a season of un- 
precedented rainfall and retarded vegetation, I could find no trace 
of Eriogonum nudum upon Catalina ; the same localities revisited 
in July, 1885, a season of early maturity, showed an abundance 
only half-grown, and it was not finally obtained in perfection 
until October, of the same year. 
Mr. Greene’s discovery of Brodia capitata in abundance on 
Guadalupe conflicts with “the almost entire absence of Liliaces”,* 
though failing to see any representative of that order upon Mr. 
Watson’s list, the “almost ” might have perhaps been altogether 
suppressed. 
That the occasional errors which appear in scientific reports 
are due to the hasty ill-digested notes of explorers is illustrated 
in the published accounts of some of these islands. ‘ 
r. Cooper found San Clemente to be ‘‘an island with 
Scarcely any soil covering the rocks” * * * * * “and 
Seems never to have been much resorted to by animals.” * The 
first of these propositions is true only of the immediate neighbor- 
hood of the usual landin , which is environed by sterile rocks, 
and at low points along the coast by long reaches of barren sands. 
he mesas or table lands of the interior, however, show a great ex- 
tent (many thousands of acres) of fine organic soil of great depth 
and apparent unbounded fertility. The second proposition is 
almost as faulty ; of marine mammals such as seals, sea lions, ete., 
it has always, until exterminated, been the favored resort ; while 
the island, since the earliest settlement of the country, has been 
Proe. Am. Acad. 1. ¢. p. 3. 
*Geology of California.—Vol. I. p- 183. 
