12 8 Memoir Sears Foundation for Marine Research 



chamber of the three flights, where the water, at least part of the time, is nearly fresh. 

 No evidence of its entrance into the strictly fresh water of Gatun Lake was obtained; 

 it seems improbable, therefore, that it uses the locks as a passageway; more probably 

 the locks serve merely as a feeding ground. If it migrates to sea to spawn, as might be 

 supposed since its young pass through a leptocephalus stage, the location of its 

 spawning ground (or grounds) remains unknown. 



Food. Its diet has not been thoroughly studied, but the short alimentary canal 

 and the well-armored mouth surely suggest a flesh diet. A dozen or so adults taken 

 in one catch at Beaufort, North Carolina, had fed exclusively on shrimp, and other 

 examples from the same vicinity had fed on them also {48: 352; 64'. 116). It has 

 been reported that an individual from Haiti had eaten a shrimp and two small fish of 

 its own species (5: 33). A specimen taken in Connecticut had fed on two small fish, 

 probably Menidia (51: 113). According to Kaplan, the Ladyfish feeds on shrimp, 

 shiners, and squid (^6: 91). 



Parasites. The cestode parasite, Rhynchobothrium bulbifer Linton, has been found 

 in the viscera of the adult Ladyfish {48: 352). 



Relationships. This species is closely related to E. affinis Regan of the Pacific coast 

 of tropical America, apparently differing only in the number of gill rakers: in 41 spec- 

 imens of E. saurus examined, 9 had 6 rakers, 16 had 7, and 16 had 8, exclusive of 

 rudiments, on the upper limb of the first arch; and in 88 specimens, 2 had 10 rakers, 

 7 had 11,13 had 12, 28 had 13, 35 had 14, and 3 had 15 rakers, exclusive of rudiments, 

 on the lower limb. In nine specimens of E. affinis examined, 4 had 10 and 5 had 11 

 rakers on the upper limb, and in 26 specimens, 2 had 16 rakers, 4 had 17, 6 had 18, 

 I I had 19, and 3 had 20 rakers, exclusive of rudiments, on the lower limb. 



Elops saurus has been recorded for West Africa by Fowler (25: 155) and by 

 others, but Regan {58: 38) has held that African specimens were distinct on the basis 

 of fewer vertebrae — 68 or 69 compared to 78 or 79 for American specimens — and 

 on the basis of fewer scales — 94—98 for African specimens and 102— 1 18 for American 

 specimens. According to counts made by me, these differences seem to be constant. 

 Ten American specimens have 73, 74, 75, 75, 75, 77, 78, 79, 80, and 80 vertebrae, 

 and nine American leptocephali have 77, 78, 78, 78, 79, 80, 82, 82, and 82 myomeres, 

 whereas one adult from Ashantee, Africa, has only 67 vertebrae. Counts of scales in 

 22 American specimens resulted as follows: i specimen with 103, 2 with 105, i with 

 106, 2 with 107, 2 with 108, I with 109, 2 with iio, i with iii, 2 with 112, i with 

 113, I with 114, I with 115, 2 with 116, i with 118, and 2 with 120. The three 

 specimens from Ashantee, Africa, examined by me have 92, 97, and 100 scales. 

 Although the range of counts for American specimens is rather great, it does not 

 overlap that given by Regan for African specimens. On the basis of the above data 

 given by Regan and me, the African representatives, which Regan named Elops senega- 

 lensis, may still be considered distinct, though synonomized with E. saurus by Fowler. 



Variations. No geographical variation in the number of gill rakers has been dis- 

 covered, but there appears to be a slight increase in number with age, which seemingly 



