162 U.S. NATIONAL MUSEUM BULLETIN 292 



upper margin. Upper surfaces of remaining pereiopods orange tan 

 with white spots and piu-pHsh-black markings; meri conspicuously- 

 marked with transverse dark Unes and large white spots, particularly- 

 abundant on second, third, and fourth pereiopods; white spots on 

 merus of fifth largely confined to marginal area; distal podomeres 

 with linear arrangement of purplish-black spots at bases of setal 

 tufts. Lower surface of legs orange-red anteriorly fading rapidly 

 posteriorly thi'ough yellow, with most of lower sm-face of basal 

 podomeres white; those distal to merus mostly orange to yellow; 

 lower distal area of meri lavender to mauve. Sternum white with 

 median purple band having undulating and angular lateral borders. 

 Abdomen mostly white; basal somites lavender and remaining ones 

 and telson with paired pale lavender markings. 



Material examined. — The Dominican collections contain 7 males 

 (carapace lengths 15.2-35.2 mm) and 3 females (cl 19.0-30.4 mm). 



Ecological notes. — Goniopsis cruentata was observed at only two 

 localities on Dominica. The fu'st place it was seen was in a low, some- 

 what muddy area near the mouth of the Mero River, where one 

 crab was standing motionless by the side of the stump of a felled 

 tree. Some few minutes were spent in chasing it from one side of the 

 stump to another until it sought refuge in a hole, perhaps its own, 

 that descended beneath the stump. Repeated visits to the same 

 stump and the area for several weeks failed to reveal this or other 

 members of the species. Burrows of Cardisoma gumihumi and Uca 

 burgersi were numerous in the area, and, at almost any time, scores 

 of the former coidd be seen at or near the mouths of their burrows, 

 but no Goniopsis was evident. 



At a marsh adjacent to the Indian River at Portsmouth (pi. 4a), 

 a number of individuals were seen along its margin. Here, on being 

 disturbed, they scurried into holes or ran out onto the marsh, where 

 they were lost from view among the bordering aeroids and lianas. They 

 would permit a person to approach them within 6 to 10 feet if the ap- 

 proach was a slow one, and it was found that a "squht" of formalin 

 directed upon them (a plastic "squeeze bottle" was used) would cause 

 them to rush into a nearby burrow; shortly they would reappear at 

 the surface — more quickly if additional formalin was poured into the 

 burrow. After such treatment, they were much more easily caught; 

 even so, many of them came out of the burrows that they had entered 

 at such a rate of speed that it was exceedingly difficult to grasp them, 

 and most rushed out onto the marsh or disappeared into another hole 

 close by. Most of the burrows that they were seen to enter were 

 horizontal ones close to Pterocarpus or under stumps or large trunks 

 of trees partially embedded in the mud. Several young Goniopsis were 

 seen running about in the arum (Montrichardia arborescens) thickets. 



