DECAPOD CRUSTACEANS OF THE WEST INDIES 87 



differences, as suspected by Pocock and others. Contrary to the 

 statement made by that author, the rostrum of many shrimps in- 

 creases rapidly in relative length in the youngest juveniles, then 

 gradually decreases in proportion as the body lengthens and broadens. 

 This pattern is followed very closely in Xiphoearis (fig. 18). In the 

 collections made on Dominica in 1964, the rostrum-carapace ratio 

 usually varies from 1.0 to 1.3 in the smallest juveniles with a cara- 

 pace length of 2.0 mm. The highest ratio, i.e., the greatest propor- 

 tionate rostral length, varies from about 1.7 to 2.2 and is attained at 

 a carapace length of about 2.3 mm. From that maximum the ratio 

 decreases rather rapidly and regularly to a generally observed range 

 of 0.8 to 1.3 in mature shrimps with carapace lengths of 8 to 12 mm. 

 With one exception (a male in which the rostrum may have regener- 

 ated), the ratios in the largest specimens taken in 1964 did not fall 

 below 0.8, and the rostrum is longer than the carapace in most of 

 the specimens, which are therefore referable to typical X> elongata as 

 restricted by Pocock; in only five specimens does the rostrum fall 

 short of the end of the antennal scale, as m Pocock's X. gladiator 

 gladiator. Much to our surprise, therefore, specimens collected in 

 Mannet's Gutter (station 50) in 1966 have the rostrum rather con- 

 sistently shorter than those taken at the same season of the year 

 from the same locality two years earlier. Seven of the 12 specimens 

 with carapace lengths greater than 8.0 mm taken in 1966 have rostrum- 

 carapace ratios as low as 0.47 to 0.55, which would refer them to or 

 near Pocock's X. brevirostris. We are at a loss to explain this striking 

 difference between the Mannet's Gutter populations of 1964 and 1966. 

 The only ovigerous female in the collection was found in Mannet's 

 Gutter on March 21, 1964, and other specimens with eggs were seen 

 but not collected at that time. This was the latest date on which 

 mature specimens were collected, and the absence of egg-bearing 

 specimens among the numerous mature females collected earlier would 

 suggest a breeding season starting in late March. On the other hand, 

 the innumerable small juvenile specimens schoohng in the lower 

 reaches of the Layou on February 12 would suggest a hatch in late 

 fall or early winter and, therefore, an extended or multiple breeding 

 season, unless the eggs are retained for a very long period or the larvae 

 are unusually long lived. 



Family Palaemonidae 



Subfamily Palaemoninae 



Key to the Species 



1. Carapace with two spines on anterolateral margin, antennal and branchio- 

 stegal Palaemon (Palaemon) pandaliformis (p. Ill) 



