Prototlieca of ilie Madreporarla. 15 



modifications can still be easily seen ; but in others tliey arc 

 at once obscured, incorporated perhaps in the subsequent 

 stock, or, ao-ain, in others worn or dissolved off. 



The tendency has been to regard the variations at the 

 extreme bases of those Palaeozoic corals as accidental, and hence 

 of no real value in classification. This view will, I hope, for 

 the future be abandoned and special attention be paid to any 

 traces which can be seen of tlie different ways in which the 

 early prototheca was modified. It is quite possible, indeed, 

 that many will be found to have been largely accidental. For 

 instance, such a bend over as that shown in diagram fig. 9 

 may have been pure accident. The same may be said of 

 radicle formation. More extensive comparisons, especially 

 from this point of view, are necessary before we can say 

 whether such a method of formino; a broad base as that shown 

 in fig. 9 became habitual in any group of early corals or not. 

 It is worth noting that other corals are known which adopted 

 it, as, for instance, the Dipterojyhylluin glans of Roenier 

 (' Lctha^a Geognostica,' i. p. 371). 



More interesting, however, than these irregular, one-sided 

 bendings over are those which took place more or less symme- 

 trically all round. The most perfect of these methods, and, I 

 believe, one of the most recent, is certainly that in which tlie 

 edge of the prototheca is very early bent down, that is before the 

 cup has any real depth, as already described above (see p. 5) 

 as being the case in the Perforata. The successive bendings 

 down and attempts to bend up again of the edge of this proto- 

 theca will, as we have seen, account for the successive 

 wrinkling of the flattened epitheca (see diagram fig. 10). The 

 Perforata owe their leading characteristics to this fact, that 

 upon their flattened prototheca or epitheca a purely septal 

 theca arises, and as the polyps bud the new thecae are also 

 septal and may mount upwards to form enormous stocks 

 built entirely out of radial septa mutually supported by con- 

 centric synaptacula3, leaving the ej)itheca, as in Turbinaria^ 

 as a film beneath the base of the stalk. 



On the solution of the question as to when this very early 

 flattening out of the prototheca arose depends that of tlie first 

 appearance of the Perforata in the coral system. We get 

 what appears to be a flat, very wrinkled epitheca in Cyclolites 

 of the Secondary epoch, and again still earlier in the Palseozoic 

 Palceocyclus. But an examination of specimens of these at 

 once shows that their flattened epithecae were not continuous as 

 in fig. 10. In Palceocyclus the conditions may be represented 

 by the diagram fig. 11. and for Cychlites by diagram fig. 12, 



