In P . monostichus (see Text-figure X), in P. oligopodus and in P. ezstuari, P. Johnsoni 

 and Botryanttms benedeni the arrangement is essentially the same. On either side of the short, 

 sterile directives are two sterile mesenteries, both of which are considerably longer than the 

 directives, though but little longer than the sterile deuterocnemes, those of the second couple 

 being longer than those of the third. The fourth couple of protocnemes consists of fertile 

 mesenteries which reach the aboral pole. 



In four distinct species, then, the telocnemes are the fourth couple counting from the 

 mid-ventral line, and this arrangement seems therefore to have more than mere specific value. 

 If so, the question arises as to whether indications of the two arrangements can be found in 

 the early stages of development. A study of the various larval forms described so thoroughly 

 by van Beneden (1898), together with my own observations on some additional ones obtained 

 by the "Siboga", have led to some interesting results in this connection. 



Van Beneden has divided the larvse he studied into two groups characterized by the 

 appendages which occur on certain of the longer mesenteries. In one group these mesenteries 

 bear peculiar aggregations of cells which may be set free within the coelenteron and have been 

 termed botrucnida, while in the other group the appendage is a more or less filamentous 

 structure which van Beneden has identified as an acontium. The Botrucnidiferse need not con- 

 cern us here; but in the Acontiferse will be found certain variations in the mesenteries which 

 bear the acontia, and in correlation with these apparently certain other peculiarities of struc- 

 ture. Unfortunately the acontia are in some cases a little late in making their appearance, and 

 some of the larvse obtained do not possess them, but in several of such cases one or more 

 of the other peculiarities are evident and from these it may be predicted upon what mesenteries 

 the acontia will appear. 



To consider first the acontia. These occur as a rule only on one or two couples ot 

 mesenteries, the more ventral of which may be either the second or the fourth protocneme. 

 In Arachnactis albida, in his three species of Ovactis and in Dactylactis armata and D. digitata 

 van Beneden found the acontia upon the fourth couple and they are also clearly shown on 

 the fourth couple in Boveri's fig. 3 (1890) although he did not perceive their significance ; 

 Gravier (1904), found them on the fourth and sixth couples in D. benedeni ; and to this list 

 I can add Ovactis wilsoni and from the "Siboga" material Arachnactis siboga and Dactylactis 

 malayensis, all three with the acontia on the fourth couple, so that it would seem fairly certain 

 that in these three genera, Dactylactis, Arachnactis and Ovactis, the acontia are borne by the 

 fourth protocnemes. In the species of the three remaining genera described by van Beneden 

 the acontia had not appeared in the larvae studied, but in the "Siboga" collection I find 

 representatives of two of these genera, namely Apiactis and Peponactis, and in the species 

 belonging to the former the acontia are borne by the second protocnemes and in the species 

 of Peponactis they are present on both the second and fourth protocnemes. 



In the more advanced larvse of the first group it is found that in addition to the fourth 

 couple being the first to possess acontia, the are also the longest of the protocnemes. This is 

 the case in Arachnactis albida, A. siboga, Ovactis brasiliensis, O. cequatorialis , O. wilsoni, 

 Dactylactis armata, D. digitata, D. benedeni and D. siboga. Of the other known species of 



