190 THE APODID^E PART n 



disposed to think that it had entirely disappeared. 

 In Apus it has not yet quite disappeared, but it is so 

 rudimentary that it appears to be in the act of disap- 

 pearing. It seemed to us that the condition of this 

 limb in Apus helped us in pointing out a missing limb 

 in Limulus. But maturer thought led to the homolo- 

 gising of the second pair of limbs with the second 

 antennae of Apus. In the first place the position 

 agrees very well (cf. Fig. 43 with Frontispiece). 

 This was not, however, the real reason for our change 

 of opinion, which was due to a comparison of the sixth 

 limb in Apus with the sixth limb in Limulus, and, to 

 anticipate somewhat, with the sixth limb in the Eury- 

 pteridae and in the Trilobites. The sudden specialisa- 

 tion of this limb in all these animals must be admitted 

 to have some common significance. That given on 

 p. 44 seems the most probable, viz., that, taking five 

 segments to form the bent head, the sixth was the 

 first free segment, and its parapodium was thus free 

 to develop into a limb for locomotion or for some 

 other function. Thus, taking the sixth limb of 

 Limulus to represent the first trunk limb, the full 

 number (five) of head limbs were left to be homologised 

 with the typical head limbs of Apus and the other 

 Crustacea. The second antenna is therefore present, 

 exactly corresponding in position with the homologous 

 limb of Apus. It is, however, a well-developed and 

 highly functional limb. In the female it is chelate, 

 but in the male it develops a seizing hook. It is in- 

 teresting to note that the same pair of limbs in the 

 male of Branchipus, which is closely related to Apus, 



