262 BULLETIN OF THE 



Near the base the dactylus divides into two branches, a long one (a), 

 which appears to be the distal part of the original dactylus bent so as to 

 make almost a right .angle with its proximal portion, and a shorter one 

 which forks at the end (6, c), and presents a row of teeth on both the 

 inner and outer borders. This shorter branch has the normal direction 

 of the dactylus, and is probably a secondary outgrowth from the primi- 

 tive dactylus. This malformation resembles that seen in Plate I. figs. 

 7 and 8. The propodite is lost. 



Plate II. Fig. 5 (Callinectes hastatus, left chela). — The dactylus is 

 divided longitudinally, nearly to its base, and furthermore the lower of 

 the two branches thus produced forks at a point midway between the 

 base and the tip. One of the prongs of the fork (c) inclines toward the 

 upper branch of the dactylus (6), the other prong (a) is curved down- 

 ward toward the index finger. The dactylus thus becomes tridactyle 

 instead of monodactyle. The superior branch (b) is toothed along its 

 lower edge, the inferior branch along both its upper and lower edges, 

 the teeth of the upper edge being continued along the upper margin of 

 the upper prong (c), while the teeth of the lower edge are continued 

 along the lower margin of the lower prong (a). All the branches are 

 much shorter than the index finger. The teeth on a do not strike 

 against those on the index when the claw is shut. Even the coloration 

 of a, b, and c is like that of the normal fingers. 



This monstrosity is like that described and figured by Lucas (Carcinus 

 incenas, in No. 7 of the Bibliography). I differ from Lucas in the inter- 

 pretation of the finger-like parts of the tridactyle segment. He con- 

 siders b to be the normal dactylus, and a and c to be supernumerary 

 fingers, a being the analogue of the dactylus (6), and c the analogue of 

 the index. From the analogy of this deformity with those represented 

 on Plate I. figs. 1-5, I conceive a to represent the original dactylus, and 

 b and c to be the supernumerary parts, representing the dactylus (a) 

 and the index respectively. 



Plate II. Fig. 6 {Romarus Amei-icamis, right cheliped). — The first 

 segment (coxa) is wanting. The second and third segments, instead of 

 having their normal flattened form, are subcylindrical. The third seg- 

 ment (meros) further shows a tendency to divide, a deep groove running 

 across the distal end. The upper half of this segment repeats antitropi- 

 cally, or in a reverse manner, the lower half : thus the spine sp on the 

 anterior border is symmetrically repeated in sp', and the articulating 

 process z has its homotype in z'. The symmetry of the segment is not 

 complete, however, inasmuch as the two or three short spines on the 



