REPORT ON THE PYCNOGONIDA. 117 



Bell, e.g., the commissures between the third and fourth ganglia are the longest ; in 

 Nymplion hrachyrhijnchus, however, those between the second and third are the longest. 

 In Colossendeis and in Phoxichilidium inlosum the relative length is the same as in 

 Nym'phon rohustum. In all the species and genei'a I observed the commissures between 

 the last two thoracic ganglia to be shorter, often much shorter, than those between the 

 preceding ganglia. The nerves which arise from these ganglia in Colossendeis immediately 

 divide into two branches, whereas in Nymplion they only separate after having reached 

 the middle of the lateral process at the end of which the leg is inserted. In Phoxi- 

 chilidium pilosum they again divide immediately after leaving the ganglion. The last 

 ganghon has a truncated form ; in most of the genera I studied I observed at least a 

 trace of the presence of the small accessory ganglia (abdominal ganglia) which Dohrn 

 observed in Phoxichiliis, and which I found beautifully developed in a young specimen 

 of Colossendeis proboscidea. 



Close to the hinder margin of the ganglion two smaU excrescences arise from its 

 dorsal surface. They are separated from one another hy a small space, and as they unite 

 again at a short distance from the ganglion a round area is left between them. In a 

 full-grown specimen of the same species only a single excrescence was observed, arising 

 from the hinder margin, and I obsei"ved it of the same form in Colosseiideis megcdonyx 

 and also in C. leptorhynchzis. I did not oljserve it in Nymphon; but in the species 

 of that genus which I studied (iV. strbmii and N. rohustum) I always saw two distinct 

 medullary nuclei in the interior of the ganglion placed close to its hinder margin, behind 

 the large medullary masses from which the nerves for the hindmost pair- of legs arise. 



According to Dohrn, the nerves which arise from this ganglion are two in number, 

 besides the two stout nerves for the hindmost pair of legs. It is extremely difficult to 

 avoid confounding threads of connective tissue, arising from the neurilemma that surrounds 

 this ganglion with nerves. I have, however, distinctly oljserved that often four different 

 nerves branch of from the ganglion ; in Colossendeis prohoscidea, e.g., two smaller ones 

 are given off more in the middle from the dorsal excrescence, and two stouter ones 

 leave the ganglion nearer to the nerves which innervate the fourth pair of legs. Two 

 pairs of nerves arising from the hinder surface are also present in G. leptorhynchus ; but, 

 strange enough, only one pair (that arising from the excrescence in the middle of the 

 hinder margin) in Colossendeis megalo7iyx. In the genus Nymphon the two nerves which 

 arise from the ganglion leave it as one single nerve, and only afterwards divide into two 

 branches. The same difficulty is experienced in ascertaining the presence of nerves 

 branching from the commissures which unite the different thoracic ganglia. These are 

 present, however, in all the genera, and even in all the species I studied. As a rule, they 

 leave the commissures much nearer to the preceding than to the following ganglion, but 

 often also nearly in the middle of the commissures; between the second and third, 

 between the third and fourth, and between the fourth and fifth thoracic ganglion, 



