Afadreporan'a?} Genus Turbinaria. 517 



under comparison. The average sizes of tlie calicles may be 

 taken and the average number of septa. But certain other 

 characters appear to me to be of even greater importance. 



1. Characters of the Protuberant Calicles. — I do not here refer 

 to the fact that the calicles may or may not protrude, but, lohen 

 they do protrude, to the way in wliich this takes place. PI. XX. 

 fig. 3 shows three different ways in which the calicles may raise 

 up the ccenenchyma — the conical, globose, and cylindrical 

 protuberances. Great as is the variation in the degree of pro- 

 tuberance found in one and the same specimen, the character of 

 the protuberance, when it does occur, is apparently constant. 

 Of course these three types are subject to an enormous number 

 of variations ; but every form approaches one or other of these 

 three. 



2. The Character of the Septa. — I have not found dentition 

 or granulation of the septa of much value, although perhaps 

 they should be taken into account (see below on the ccenen- 

 chyma). Of greater value is the relation between the cycles 

 of septa and the polyp-cavity. I have found it useful to 

 draw an imaginary half -radius circle within the aperture of 

 the calicle (PI. XX. fig. 4, b). The septa may (a) fall short 

 of this, leaving a large central fossa (fig. 4, 1) ; {b) they may 

 reach it, leaving a medium-sized central fossa (fig. 4, 2) • 

 (c) they may cross it, in which case the central fossa is very 

 minute (fig. 4, 3) ; or (d) they may be quite irregular, some 

 crossing, others not even reaching it. 



Again, the septa in thus projecting into the cavity may run 

 in on a level with the margin and then dip suddenly down 

 towards the columella, or they may curve regularly round, or 

 they may slope down gradually so that the central fossa is 

 funnel-shaped. 



We have accordingly many different sizes and forms of 

 fossa?. 



3. The Interseptal Loculi. — These are also of importance 

 (PI. XX. fig. 4) ; they may be large and open or narrow and 

 slit-like. They may be almost square or petaloid, i. e. with 

 neatly rounded peripheral margins ; they may, indeed, have 

 no distinct ])eriplieral margins, i.e. the interseptal space runs 

 on continuously witii the surface-furrows of the ccenenchyma. 

 When this is hequent it is an index of the large size of the 

 pores connecting the polyp-cavities between the septa with 

 the canal-system of tlie ccenenchyma. 



4. The columella offers characters of value. 



Important as these characters undoubtedly are, it has again 

 to be pointed out that they arc only strictly applicable to 



