NATURE OF THE STREAM 131 



vice as to the quickest way in which a vessel may steam south 

 against the current with the least delay: ''A steamer bound 

 from Cape Hatteras to Havana or the Gulf ports crosses the 

 Stream off Cape Hatteras. A fair allowance to make in crossing 

 the Stream at right angles is IV2 knots per hour [sic] for a 

 vessel's speed of 5 knots for a distance of 40 miles from the 

 100-fathom curve. In the run from the southern edge of the 

 Stream to Matanilla Shoal, no allowance for current can be 

 given. Upon sighting the Bahama Bank, time will be saved by 

 running down the Stream on the east side as far as Gun Cay 

 instead of crossing at Jupiter and running the latitude down 

 on the Florida side of the channel. The current is weak on the 

 Bahama side, and on the shoals there is practically none. This 

 route will be difficult and perhaps impracticable until a light- 

 house is built at Matanilla, unless the green water of the 

 northwest corner is sighted before dark. Arriving at Gun Cay, 

 Bahama, an allowance of IV2 knots per hour for speed of the 

 vessel of 5 knots per hour will make a course of west good to 

 Fowey Rocks. This is the average velocity of the Stream. The 

 weakest current will be experienced about three hours before 

 the transit of the Moon, and if the crossing is made so as to 

 arrive at the axis at about this hour, time will be saved.'' Pills- 

 bury advises the navigator of a vessel steaming north to take 

 advantage of the periodic shift in the River's axis: ''At high 

 declination he can edge out so as to pass Fowey Rocks light- 

 house 7 miles distant, and be sure of a good current, while, at 

 low declination the maximum velocity at this distance will be 

 found much less, and it will be necessary to go 4 or 5 miles 

 farther to the eastward." 



The Blake investigations have deservedly been called a 

 classic in oceanographic current studies. No similar large-scale 

 survey has been made until recent years, but the work of col- 

 lecting and summarizing data from the observations of numer- 



