and Affinities of Graptolites. 371 



canal of the perisarc (see fig. 3) after the destmction of all 

 the soft parts is entirely similar to the mode of communication 

 between the calicles and the common canal in the fossil (in 

 those cases, at least, in which the graptolite has afforded faci- 

 lities for examination such as to leave no doubt as to the struc- 

 ture of the parts in question) , and quite different from that in 

 which the proximal extremity of the hydrotheca is connected 

 with the common tube of the chitinous perisarc in the existing 

 hydroid*. 



I cannot help believing that this is the true view to take of 

 the morphology of graptolites. If so, the graptolites would 

 admit of an approximation through an unexpected channel with 

 the Pluraularidaj. They would then be morphologically plu- 

 mularians in which the development of hydrothecaj had been 

 suppressed by the great development of the nematophores, 

 probably the mesial onesf ; while, on the other hand, the exist- 

 ing plumularian with well-developed hydrothecse would pre- 

 sent in its nematophores the last traces of the structure of its 

 ancient representative, the graptolite. 



That the complete suppression of the hydrotheca3 simulta- 

 neously with the retention of the nematophores is no over- 

 strained supposition, will be admitted from what may be seen 

 in certain plumularian hydroids which carry peculiar branches 

 destined for the support of the gonangia or generative cap- 

 sules. Now these branches are always destitute of hydro- 

 thecae ; but they are richly supplied Avith nematophores, which 

 are distributed along the length of the branch, sometimes in a 

 single row like the denticles of the monoprionidian graptolites, 

 and sometimes in two opposite rows, like those of the diprio- 

 nidian forms. In one undescribed species, from the deep-sea 

 dredgings of the ' Porcupine,' I have found quite similar 

 branches sent off from parts where they can have no connexion 

 with the generative functions of the colony. The resemblance 



* In the older parts of tlie hydroid stem the chitinous walls may be- 

 come much thickened by successive layers of chitine, and the communi- 

 cation between the common canal and the cavity of the nematophore 

 may thereby become contraeted — a condition, however, which must not 

 be confounded with the nature of the communication between the hydro- 

 theca and its supporting stem. 



t It may be here suggested that while the calicles of the monoprionidian 

 graptolites have their representatives in the azygous or mesial nemato- 

 phores of the plumularian, those of the diprionidian graptolites are repre- 

 sented by the paired or lateral nematophores. I should not hesitate to 

 maintain this view, were it not that the comparison of a pair of opposite 

 calicles in a diprionidian graptolite with a pair of lateral nematophores in 

 a plumularian could hardly be reconciled wdth the view which \yould 

 (apparently with reason) regard the diprionidian forms as morphologically 

 representing two monoprionidian forms united back to back. 



