CYCLOPOIDEA. 1027 



length ; and in some males a like difference exists between the left 

 and right antennae. Among the genera with long antennae, therefore, 

 where the number of joints is few, it is so from a union of several 

 joints that are in other cases separate. Figures 37, 38 belong to the 

 same species ; in the former several joints correspond to each of the 

 longer ones in the latter. So in figures 18 a and b, 19 and 20, 26 

 and 27, 28 and 29, 30 and 31, 32 and 33, each of the longer joints 

 of one antenna, in each species, corresponds to several joints in the 

 other. 



This is very evident from the occasional existence of the articu- 

 lations, either complete or in faint traces, intersecting the longer joints. 

 In figures 30, 31, the last three joints of 31 correspond to the last 

 three of 30 ; the next of 31, to the next two of 30 ; the next (or fifth 

 from apex) of 31, to the next two (sixth and seventh from apex) of 

 30 ; the next (or sixth from apex) of 31, to at least four, probably five, 

 (eighth, ninth, tenth, eleventh, and twelfth) of 30. In figure 29, 

 these joints are actually separate, as in 28, excepting the fourth and 

 fifth from apex of 28, which are coalesced in 29 ; but the joint formed 

 has one or two setae at middle to indicate its compound character. 

 In 26 and 27, the same principle is illustrated. 



This is also obvious, from the relative positions of the setae in the 

 different antennae. For example, in figure 37, the longer setae have 

 about the same distances apart as in 38, and there is a general corre- 

 spondence in other respects, as is seen by making the comparison. It 

 is plain, therefore, that the last four joints of 38, correspond to the 

 last four of 37 ; the next of 38 to the next three of 37 ; the next of 38 

 to the next two of 37. Again, the longish seta on the broad joint of 

 figure 31, obviously corresponds to a similar one at the apex of the 

 ninth joint (from apex) of 30. Yery often, too, the second joint at 

 base is much longer in the male than in the female (as in 26 and 27), 

 and faint lines are at times distinguished that show that this length 

 arises from the non-separation of short joints from the apex of this 

 second joint. This is the case in figure 33. 



We therefore arrive at the general conclusion, that the multiplica- 

 tion of joints takes place by the subdivision of pre-existing joints. This 

 appears to be a general law, not confined to Crustacea. The increase 

 does not take place by the addition of joints at the base of the organ, 

 or their seriate development at its apex. From the first stages of the 



