10 2g CRUSTACEA. 



animal, the existing joints (of which there are but two or three in 

 the earliest state of the young) gradually lengthen and undergo sub- 

 division, and by this process the multiarticulate character is produced. 

 These changes probably take place mainly in the process of moulting. 

 Towards the base of the simple antenna, the subdivisions are near 

 the apex of the joints, and commonly from the apical half of the 

 second joint. But about the medial portions the subdivision often 

 bisects the joint. In the many-jointed antennae these medial joints 

 when much oblong have one or two setae at middle, as in figures 28 

 and 36, which setae indicate this tendency to central subdivision, 

 although the bisection does not actually take place. Upon the second 

 basal joint the setae are usually clustered towards apex, this part 

 being the portion that tends to become subdivided. Thus while 

 development takes place symmetrically about the middle and outer 

 portions of these organs, it is most active towards the outer extremity 

 of the second basal joint, instead of its middle or basal portions. 



These principles account for the fact that the first joint of the 

 antenna is nearly alike in the different species of a genus, however 

 different the length of the organs ; this would not be the case if the 

 multiplication of joints took place at the base. This first joint is, 

 however, sometimes obsolete, a fact which should be noted, since 

 ignorance of it might lead to incorrect inferences regarding homo- 

 logous parts. 



It would seem also to be true, that in the multiarticulate antennae 

 (species of Calanidae), the last three joints do not change by sub- 

 division; for these joints are furnished with posterior as well as ante- 

 rior setae (in which fact, with sometimes the next preceding joint, 

 they are peculiar), and this is the case whatever may be the number 

 of joints, which in some Calanidae is but nine. If the apical or penult 

 joint subdivided, this would increase the number of joints bearing 

 posterior setae; and the number actually found could be retained m 

 such a case only by removal of one or more pre-existing setae. This 

 removal might take place, but we have no evidence of it. It is, 

 therefore, probable, that after the organs are so far advanced as to 

 have posterior setae to the three or four terminal joints, these joints 

 do not undergo subdivision. These setae in fact appear often to pre- 

 cede the existence of separate joints for them, the separation (or 

 development of an articulation) afterwards taking place. Thus in 

 figure 40, the second joint from apex corresponds evidently to the 



