SEX RATIO IN GAMBUSIA. 393 



winter and spring than they are in midsummer. The data at hand 

 are for the following months : June, July, August, September, 

 October and December. Males were most numerous, as shown 

 by these data, in June, when the ratio was i male to 2.54 females 

 among 4,902 fish examined. In July the ratio, based on the ex- 

 amination of 17,941 fish, was i male to 5 females. In August 

 the males were the fewest, as a ratio of only i male to 11.3 fe- 

 males obtained, as shown by the examination of 21,446 fish. In 

 September the males began to increase, for a ratio of i male to 

 . 8.35 females was obtained from a total of 13,473 fish. In October 

 the males increased still further, for i male to 2.75 females was 

 found among 43,288 fish examined. A single day's collection, 

 from Beaufort, N. C., taken chiefly in brackish swamps, for De- 

 cember is at hand, which consists of 2,100 fish. The sex ratio in 

 this collection consists of i male to 3.6 females. The number 

 examined obviously is too small and the environment too limited 

 to be representative of the usual sex ratio prevailing at this season 

 of the year. The average ratio of males to females for all fish 

 (103,120) examined is i to 4.4. 



The comparatively large seasonal difference in sex ratio in 

 Ganibusia, is very interesting. Barney and Anson (1921) offer 

 the following explanation (pp. 64 to 66) : " Further inquiry into 

 the causes of the varying seasonal frequency of males and of the 

 species, shows that in the summer there is a decrease in depth and 

 area of all water systems studied. . . . This lowering of the water 

 in the bayous and ponds eliminates all the shallow margins which 

 Ganibusia frequent and where they are largely immune from their 

 enemies, and forces them into deeper water where the incline of 

 the banks is steeper. This accordingly means much less natural 

 protection than the minnows formerly had, since predaceous fish, 

 now at a period of heightened metabolism and consequent rapid 

 growth, are especially destructive and have a marked effect in 

 lowering Ganibusia frequency. Doubtless the Ganibusia eaten by 

 their predators at this time are nearly all females of large size, 

 since the adult females, gravid at this season, are much larger, 

 much slower in movement, and are more noticeable because of 

 their black abdominal spots than the small, quickly moving, uni- 

 formly colored male. With the height of Gainbifsia frequency 



