BRYOZOA g5 



In my paper on the Bryozoa from Franz Josef Land (p. 62) référence is made to the 

 Hornera lichenoides L. said to hâve been dredged by the Erebus and Terror, from Lat. 74 S.- 

 Long. 172 E., 33o fathoms on which Cribrilina punctata Hass. was growing, and with which 

 there was also Escharoidcs Sarsii Sm. The pièce of Hornera lichenoides is quite similar to Arctic 

 spécimens, and does not resemble spécimens of H. antarctica so that the opinion expressed, that 

 thèse very characteristic Arctic species were not collected in the Antarctic, but that some mistake 

 has been made in labelling, seems now much more probable, after an examination of an 

 Antarctic collection, than it did when written. The évidence seems to justify our saying that 

 a mistake has certainly been made. 



This form occurs from Cape Horn in the material brought back bv the Romanche, but 

 Julliex does not mention any similar Hornera, so that no doubt he had not seen this material. 



Habitat. — Cape Horn (W.). 



Exp. Antarct. Belge. 



N° 565, Tangles VII. Lat. 70° 23' S.- Long. 82 47' W.; 480 met.; +0.8 C. 



N os 6i3, 683, 991, Tangles VIII. Lat. 70 00' S.- Long. 8o°48' W.; 5oo ? met.; +0.9 C. 



N° 800, Tangles X. Lat. 70 i5' S.- Long. 84006' W.; 56g met.; +0.8 C. 



Suborder RECTANGULATA Waters 



In 1887 (') I proposed that the Cyclostomata should be divided into two subdivisions, the 

 Parallclata in which the surface of the zoarium is to a considérable extent, formed of the latéral 

 walls ofthe zoœcia, as in Crisia, Entalophora, Diastopora, Tubulipora. This is much the same 

 as the suborder Tubulata of Gregory, except that he recognises the Articulata, while I should 

 not attach so much importance to articulation, for although Crisia is usually articulated, there 

 are fossil forms of Crisia unarticulated. The second division was the Rectangulata in which 

 the zoœcia and cancelli open for the most part at right angles to the axis, or surface of the 

 zoarium, or subcolony, of which Heteropora and Lickenopora may be taken as types. 



Gregory has pointée! out, that this is much the same as the Order Trepostomata of 

 Ulrich; but the early stages of Lickenopora, and the larva of Lichcnopora show so much simi- 

 larity with characteristic Cyclostomata, that I should hesitate to place them in separate orders, 

 at présent. Gregory however makes a suborder Cancellata, which in his Jurassic Catalogue 

 includes the Discoporellidœ (p. 34) by this is meant the Lichenoporida , and he has previously 

 spoken of cancelli in Lickenopora (p. 12) so that hère his Cancellata would include most of the 

 Rectangulata. In the jurassic Catalogue, where he first diagnosed the Cancellata, he says p. 3g 

 « composed of simple zoœcia separated by cancelli » and goes on to say « for diagnosis see 

 part III », but there being no Cancellata to catalogue the diagnosis has apparently been omitted; 

 and the diagnosis in the Cretaceous Catalogue does not at ail correspond with that previously 

 given, for in this later catalogue (p. 35g) he speaks of « tubular zoœcia with cancellate walls», 

 which presumably would mean perforate walls. In the first diagnosis the cancelli were a 

 zoarial character, as understood by Busk and others, now they are zoœcial ; and Hornera, which 

 in the jurassic Catalogue was placée! under Tubulata, is now in the Cretaceous Catalogue 



(1) Tert. Cyclost. Bryozoa from New Zealand. (Quart. Journ. Geol. Soc, vol. XLIII, p. 337). 



