AND GLYrHA8TR.EA, DUNCAN (1887). 211 



on the same type species, Septastrcea Forbesi. In his original 

 paper, read before the Society, the presence of a columella is cate- 

 gorically asserted, and in the published Abstract* we find it men- 

 tioned as a " columella," " a narrow linear columella," and " a narrow 

 ribbon-shaped columella." In the subsequently modified and aug- 

 mented paper in the Quarterly Journal we find the same structure 

 alluded to under various terms on every page, but with the signifi- 

 cant addition that it is sometimes absent. Besides the simple 

 appellation " columella" t, it is likewise defined as " a trabecular 

 and non-essential columella, it often being reduced to a mere lamina 

 which is in the path of two opposed large primaries " J ; -'a narrow 

 discontinuous columella, which is ornamented in the same manner 

 as the septa, but Avhich, in some instances, has a raised edge " § ; "a 

 columella which sections prove to be occasionally discontinuous and 

 always non-essential in its method of growth " || ; " columella small, 

 parietal, lamellar or ribbon-shaped, uniting opposite primaries, or 

 several septa sometimes absent " (sic). 



After a careful examination of all the specimens described and 

 figured in the paper quoted above, I am unable to discover any of 

 the diff'erent varieties of columella which they are stated to possess. 

 They all may, in my opinion, be readily traced to the union and 

 partial involution of the inner margins of the septa in the axis of 

 the corallites, producing the structure known as a pseud o-columella. 

 Nor do the figures accompanying the paper (so far as they are true 

 to nature and not diagrammatic) show structures beyond this. The 

 figures given of the section made at the British Museum (pi. iii. 

 figs. 13, 14) to determine this point conclusively show that a true 

 columella is not present *f[. 



Minnie Structural Features of Septastraea. — As already mentioned, 

 the eff"ects of fossilization on the examples of Sej^tastrcea from IS^orth- 

 American Miocene strata have been the reverse of what generally 

 happens in fossil organisms ; for instead of destroying or obliteratiug 

 structural features, it has revealed some details more clearly than 

 in many recent corals, and thus thrown fresh light on the nature of 

 the thecal wall, the septa, and the so-called costae of the corallites 

 and their mutual relation to each other. 



Structure of the Septa. — On examining, under the microscope or 

 a strong simple lens, a fractured edge or transverse section of a stem 

 or branch of S. Forbesi, the individual septum of the central full- 

 grown corallites, though of great tenuity, not exceeding the thickness 



* No. 495, 1886-7, pp. 18, 19, 



t L. c. p. 24. 



t Md. p. 26. 



§ Ibid.-^. 28, pi. iii. figs. 3, 6, 9, 10. 



II Ihid. p. 28. 



^ My justification for thus entering with so much detail into the nature ot 

 the asserted columella may be best expressed in the following words of Prof. 

 Duncan respecting this structure: — "Playing a very important part in the 

 economy, and being in relation both with the septa and tlie pali, the columellas 

 are structures whose variations in form are of generic import " (Brit. Fossil 

 Corals, Suppl., Pal. Soc. 1866, pt. i. p. 14). 



