94 Dircdoi'^s Annual Report. 



be observed that the three points of the island are much higher 

 than the middle of the three sides, which is due to the points being 

 exposed to the building agencies from two direcftions. The same 

 agencies of wind and tide which have formed the island have 

 doubtless planted its coconuts, brought its vegetation and the land 

 reptiles which inhabit it. 



METEOROLOGICAL CONDITIONS. 



The winds at Marcus Island in all probability vary but little 

 from the prevailing winds in that part of the Pacific. It is well 

 known that they are of sufficient force during the greater part of 

 the year to cause the surf to break on all sides of the reef with 

 great fury, rendering landing an exceedingly dangerous, if not im- 

 possible, feat during the seven months of the year from September 

 until April. Throughout the remaining summer months the ordi- 

 nary steady trade winds blow, and it is only at this season that 

 landings can be made in safety. 



We had no means of estimating the rainfall during the year, 

 but the Japanese were able to catch enough water from the roofs 

 in barrels, jars and boats to keep them well supplied with drinking 

 water ; and as they were especially careful to store all they possibly 

 could I concluded that there were periods of drought. Aside from 

 this source of supply there was no fresh water available save what 

 little might be colle(5led in the holes in the exposed reef at the 

 north end of the island. There were frequent tropical rains during 

 our stay that were considered by the residents as out of season. 

 These caused the water to collect in the low places in the dry 

 lagoons. When examined chemically it was found to be charged 

 with phosphoric acid and organic matter to such an extent that it 

 was pronounced unfit for drinking. The colon}- had supplemented 

 the supply of water by digging a brackish water well fifteen feet 

 deep back a little distance from the shore ; this furnished water 

 suitable for cooking and washing purposes only. 



Thunder and lightning usually accompanied the storms. At 

 times the elecflrical display w-as splendid, though disheartening to 

 a naturalist. During a heavy rain on the night of August 5 there 

 occurred what Mr. Sedgwick and myself both believed to have been 

 a violent seismic disturbance. Though sleeping soundly the vibra- 

 tion caused us both to wake suddenly. We were so di.sturbed as 

 to deem it advisable to make a light ; on rising I felt it necessary 



