INVERTEBRATA, ORYPTOGAMIA, MICROSCOPY, ETC. 871 



deal, after having, with Dohrn, directed attention to the fact that the 

 investigations of the Wlirzburg professor seem to set at rest a 

 question which has been differently answered by different observers. 

 The genus PhoxicMlidimn, which has been lately subjected to examina- 

 tion, is, in its larval stages, parasitic on Hydractinia ; according to 

 Gegeubaur, the parent lays its eggs in the polyp, whereas Hodge and 

 others state that the larvfe enter the polyp so soon as they leave the 

 egg ; among the latter is Claus, and that he was right in his view is 

 shown by Semper's detailed account. 



Turning to the points on which Semper and Dohrn are at 

 variance, between PhoxicMlidimn, as described by the former, and 

 Achelia, as described by Dohrn, there is this difference, that in the 

 former the ovigerous appendages are a later development and are 

 not represented in the larva, so that the Pycnogonida apparently 

 have not as Dohrn imagines seven, but only six pairs of true appen- 

 dages. So far as this bears on the relationship of the Pycnogonida 

 with the Arachnida, which as Professor Semper thinks is rendered 

 doubtful by their i)ossession of only six pairs of legs, Dohrn remarks 

 that " the homologies of the aj)pendages will only be proved when 

 the blood-relationship of the separate orders of the Arthropoda has 

 been demonstrated by some other method." In addition to this 

 " theoretical " consideration Dohrn asserts that Semper is wrong in 

 his facts. The great and important point is the character of the 

 nervous system : with regard to the first ventral ganglion, it gives off 

 three pairs of nerves ; the first, which is not noticed by Semper, 

 supplies the rostrum and its musculature, its sensory hairs and com- 

 plicated lips ; the second pair supplies the so-called tentacles, or 

 those appendages which are intercalated between the rostrum and the 

 ovigerous appendages : the third pair pass to the ovigerous appen- 

 dages and present the same morphological relations as the second 

 pair. 



In the larva, the first pair of legs are sujiplied by a nerve given 

 off from the supra-oesophageal ganglion and the two hinder pairs by 

 two distinct nerves, which arise from two incomjjletely separated ventral 

 ganglia ; from this it is clear that these two larval extremities are not 

 equivalent to the appendages of one segment, but are two " homodyna- 

 mous structures." The larva is therefore provided with three pairs of 

 appendages, and as four other pairs are developed later on, Dohrn's 

 original view that there are seven pairs of appendages in the Pycno- 

 gonida is true for all cases in which the larval legs persist ; and even 

 when the third pair disajipear, as Semper has shown them to do in the 

 case of Phoxickilidium, the nerve still retains its old morphological 

 relations, and supplies the horseshoe-shaped ridge which is the final 

 remnant of the ovigerous appendages. From the point of view then 

 of development and of nervous relations the " typical " character of 

 the ovigerous legs (the third pair in the larva) as true appendages 

 appears to be well established. 



Professor Semper has, however, yet another argument ; he asserts 

 that in all Pycnogonida ca^cal gastric sacs are " typically" continued 

 into the appendages : of the third of these sacs he states that it 



