394 SAMUEL F. HILDEBRAND. 



in late August (1919) occurs properly the height of the young 

 Gambusia frequency. The young frequency then diminishes 

 quickly, apparently due to the scarcity of breeding males during 

 the spring and summer, and the resulting lowered birth rate. The 

 height of Gambusia frequency is immediately followed by a con- 

 siderable drop in water stage (surface water index) and accom- 

 panied by an increase in the percentage of other species, many of 

 which are predators. The Gambusia frequency decreases by loss 

 of adult females, while the male frequency accordingly increases." 

 In the light of the present studies, as well as the tables presented 

 by Barney and Anson (1921) this explanation is entirely inade- 

 quate, and in the main contrary to what actually takes place. If, 

 in the thinning out process, more females than males were elimi- 

 nated, then, certainly, there should be an abundance of males in 

 July and August. That is, at a time when few young of the cur- 

 rent season have become sexually mature and when it is usually 

 quite possible to distinguish between the young of the current sea- 

 son and those of the previous year. Barney and Anson's table, 

 as well as Table I. (presented herewith), however, show that the 

 reverse unmistakably is true. Furthermore, it may be remarked 

 here that the error in sexing, by means of external characters, very 

 probably is smaller during July and August than at any other 

 time. That is, fish of the previous season 'have become sexually 

 mature and comparatively few young of the current season have 

 reached a sufficiently large size to be considered adults. There- 

 fore, comparatively few males, at this time, are classed as females. 

 How may the seasonal variation in sex ratio of Gambusia then 

 be explained? It is shown in a subsequent section of this paper 

 that males and females are equally proportioned among young 

 Gambusia, and it already has been stated that a very large decrease 

 in the proportionate number of males present takes place shortly 

 before (July and August) the main body of young fish of the 

 current season become sexually mature. A greater thinning out 

 of adult males than of adult females, therefore, must have taken 

 place. A discussion of the evidently much more rapid decrease 

 among the males than among the females, when few young fish 

 are maturing to take their place, is reserved for a later section. 

 It appears to be sufficient to state ! here that the proportionately 



