MOVEMENTS OF VEGETATION IN THE SALTON SINK. 169 



observations on the beaches and shores of one of these, Obsidian Island, have already been 

 described. All but two or three of the numerous species which appeared on the emersed 

 strips were, in all reasonable probability, carried here from places outside of the lake, and 

 reference will be made to their action in the following pages; but some of the hills were 

 completely covered by the water at rest, or by wave-action, and these deserve special atten- 

 tion, since as experimental settings they constitute much stricter tests of the theories of 

 seed-dispersal than others in which the summits remained above the water (see Plate 31 b). 



The greater mass of the hills consisted of rock with surface layers and pockets of wind- 

 blown soil before the inundation. The action of the water resulted in washing much of 

 this soil from the hills and the saturated salt-solutions which steadily increased in concen- 

 tration may be taken to have killed or set in action all seeds present. If germination ensued 

 the seedlings would of course be floated away. 



Some of the species of seed plants which thrive in the vicinity of salt springs have 

 already been noted as surviving after a year of submergence, and a small shrubby species, 

 among the Carrizo dunes, showed similar endurance, yet none of these species was seen 

 on the islands, which appeared to be completely sterilized, except perhaps as to minute 

 organisms such as bacteria, and it seems probable that those present in the dry soil were 

 extirpated and those found after emersion are to be regarded as introductions. 



The first example of colonization of a sterilized island was met in October 1908, when 

 a visit was made to Cormorant Island, 2 miles to the northward of Obsidian Island, near 

 the location of some salses or mud volcanoes now deep under water. This island was sepa- 

 rated from the mainland by about 6 or 7 miles of water, which reached a depth of about 

 34 feet at the maximum, and a greater depth between it and the nearest island. (See 

 Plate 31 b.) The top of the island showed a rocky surface of whitened rhyolitic obsidian in 

 November 1908. No exposure of fine soil was visible and it was evident that the waves 

 had been driven across its summit repeatedly during a large part of 1907. A great number 

 of nests of aquatic birds were placed among the rocks and the surface all about them showed 

 signs of having been trampled by both young and old birds. 



Pluchea sericea and Baccharis glutinosa were each represented by one plant. As both 

 of these species are compositaceous and the fruits are carried long distances by the wind, 

 the probabilities are largely in favor of their deposition here by the action of air-currents. 

 Both species grow in abundance on the mainland 12 or 15 miles to the southward, and the 

 summer winds may have brought them here late in 1907. The seeds of both appear to 

 germinate in the presence of some proportion of salt. 



Cormorant Island lay away from the main routes of travel about the lake and its 

 rocky shores could not be gained in any kind of rough weather. The second visit to the 

 place was made in October 1912. The original individuals of Pluchea and Baccharis had 

 survived. An additional individual of each was also present. That of Pluchea was some 

 yards from the original plant, and may have arisen from one of its seeds or from a newly 

 introduced one. The second individual of Baccharis was near the original, so near that it 

 might have come from offshoots or seeds, but probably did not represent a second intro- 

 duction. That the total increase of the two species was limited to a single individual each 

 in four years was a very astonishing fact, as both of the original plants had fruited probably 

 three times and the annual crop must have included a large number of seeds, while the later 

 emersions down the slopes of the island showed many clay exposures in which germination 

 might have taken place. The island was dome-shaped, being about 1,200 feet in its long 

 axis with half that width, and rising about 30 feet above the water. The entire surface 

 was now carefully scanned, with the result that two individuals of Sesuvium and one of 

 Atriplex lentiformis were found in a place probably emersed in 1908, and one Atriplex was 

 lower down in the emersion of 1909 and one in that of 1910 and one in 1911. Heliotro- 

 pium curassavicum was represented by two individuals, both of which were in bloom, 



