SPONGES, GRAPTOLITES, CORALS. 73 



AnomalosponRla.] 



constriction of the vertical ray immediately beneath the horizontal rays. This is 

 relatively greater in Anomalospongia than in any other form known to me. 



To resume: we have among the differences ( 1 ) the total absence of summit 

 plates, (2) three instead of four horizontal spicularrays, (3) the duplex character, and 

 (4) the interweaving of the horizontal rays, (5) the contact between the club-shaped 

 vertical rays, and (6) the uniform size and different arrangement of the spicules. 

 Opposed to these we have as points of agreement, ( 1 j the form and comparatively 

 large size of the vertical or entering ray, (2) its arrangement in the sponge-wall per- 

 pendicular to the surface, and (3) the possession of relatively small horizontal rays. 



This concise statement of the points of likeness and of difference is I believe 

 sufficient to show that Anomalospongia cannot be placed in the same family with 

 ReceptacuUtes. Still. I am satisfied that real relationship, however remote, exists 

 between them. As I now view the matter it seems advisable to introduce a new 

 order for the reception of the Receptaculitida>, Anomalospongia, and also Amphispongia, 

 Salter ; the relations between the last two seeming to be, as I will endeavor to show 

 pi'eseutly, closer than might be suspected from a casuat comparison. 



The new order would be strictly paleozoic, and, excepting a few forms that sur- 

 vived into the Devonian and possibly later, would be essentially Silurian. It would 

 therefore comprise only early types that, in common with nearly every class of 

 animals represented in paleozoic times, may be called comprehensive because they 

 combine characters which in more recent times became separately developed and 

 diagnostic of now widely different groups of genera and families. Perhaps the most 

 striking diversity in these respects, shown by the forms in question, is the difference 

 in the number of horizontal rays pertaining on the one hand to the Receptaculitidcc, 

 with four, and on the other to Anomalospongia with three. 



In the number and disposition of their rays the spicules of Anomalospongia 

 remind us of true Tetradinellidce. They also resemble, perhaps even more, the trifid 

 surface spicules ("Gabel-Anker'') of many lithistid sponges. The horizontal rays in 

 the latter often are bifurcate close to the centre, so that even the duplex character 

 of these rays in Anomalospongia is in a measure simulated. (See fig. \b and c.) I 

 am not prepared to decide definitely that these resemblances are or are not indica- 

 tive of I'elationship. It seemed desirable, however, to mention the facts, since they 

 illustrate the sense of the preceding paragraph. 



As already indicated, it is my belief that the uncertain Amphispongia is related 

 to A nomalospongia indeed, that the two might well be united in one family. That 

 genus was proposed by Salter* for certain free, compressed, elongate-elliptical masses, 

 rounded at both ends, and rarely more than 50 mm. long by 18 mm. wide, which 



*Mem. Gcol. Sur. Gt, Britain. 32. Scotland, p. 135, 1S61. 



