MECCA TO FIGTREE JOHN 179 



practically no supplies of water, tended to disappear 

 by evaporation. 



From time to time, however, the river must have 

 broken in, with the result that the lake became 

 brackish. Thus, the shells that are a noticeable fea- 

 ture of all the below-sea-level area are of kinds 

 native to fresh or brackish waters. The shell-remains 

 of the original sea epoch are now found high above 

 sea-level, betokening some great upheaval in remote 

 times. It is to the brackish period that the deposits 

 of travertine (calcium carbonate) are due. Proof can 

 be seen, in marks of old lake beaches at various 

 levels, that there was a succession of complete or 

 partial fillings and emptyings of the lake basin, the 

 inflow no doubt usually coming from the river, but 

 perhaps sometimes from the Gulf. 



From Indian tradition it would seem that for a 

 long time prior to recent years the lake bed as a rule 

 has been dry. Great deposits of salt occupied the 

 deepest portion, and, a few years ago, were being 

 worked on a large scale. In 1891 there occurred a 

 relatively small inflow from the river, creating a 

 shallow lake of some two hundred square miles : but 

 in 1905, through the weakness of levees and head- 

 gates of the canal system that was carrying the 

 Colorado River water on to the lands of the new 

 Imperial Valley settlement, came a greater flood 

 which caused serious loss and threatened a whole- 

 sale disaster. For over two years the water rose, 

 until it seemed as if it would entirely fill its old basin. 

 It was not till early in 1907 that the Engineer finally 

 conquered the River. (I say finally, but after all that 



