NATURE'S OWN SEAPLANES HUBBS 343 



modified by a strong wind. They tilt their plane surface more or 

 less upward toward the wind and are thus carried olf their original 

 courses in a wide curve. To what degree the wind is so utilized to 

 prolong or definitely alter the direction of the air journey is a point 

 worthy of detailed inquiry. 



THE DURATION OF THE FLIGHTS 



A person's offhand estimate of the time a flying fish remains in 

 the air is usually several times too high. During my observations 

 on the species of biplane flying fishes in Asiatic waters I timed 424 

 flights or parts of flights. The two longest single flights, during 

 which the flsh remained completely out of the water, lasted only 12 

 and 13 seconds, respectively. Some flights which I observed among 

 the Philippines, before I started to use the watch, seemed to be some- 

 what longer than any of those actually timed. Of 42 compound 

 flights that were timed the longest lasted slightly less than 30 

 seconds. It is doubtful whether any flights cover a whole minute. 

 The record flight actually timed is perhaps one of 42 seconds, re- 

 corded by Breder (1929) on the authority of a sea captain. 



Most flights are short, lasting only 1 or 2 seconds, whether the sea 

 be glassy smooth or decidedly rough. The average time for the 

 first flight was 2.7 seconds for the smooth Bohol Strait, and 2.9 

 seconds for the rough weather on the open ocean. The average 

 time for compound flights was only 2.6 seconds. 



THE NUMBER OF SUCCESSIVE LEAPS 



In order to obtain some definite data on the frequency and ex- 

 tent of these compound flights which characterize the air move- 

 ments of the biplane flying fishes, I made numerous counts of the 

 number of successive flights. Slightly more than half of the flights 

 counted (164 out of 299, or 55 percent) were not continued at all. 

 About three-fourths (76 percent) ended with either 1 or 2 leaps. 

 Nearly nine-tenths (86 percent) of the total ended with 1 to 3 

 leaps, and more than nine-tenths (94 percent) ended with 4 or fewer 

 stages. Only 1 flight in 16, on the average, was made up of 5 or more 

 elements. The greatest number of successive flights observed was 

 12. It is doubtful whether a greater number is often undertaken. 

 The numbers of successive flights under the different conditions of 

 sea, weather, size, and species of fish was found to be independent of 

 these conditions. 



THE LENGTH OF THE FLIGHT 



How far flying fishes fly apparently remains to be measured. 

 Simultaneous use of two sectants, or the combined use of a geologist's 

 compass and a distance finder, would seem to be a feasible way to 



