C. E. Gordon— Early Stages in Paleozoic Corals. 113 



Figure 3 represents "a transverse section of a third corallum 

 taken a little higher than in the previous sections. Duerden 

 remarks that the six primary septa are recognizable by their 

 greater size. The outlines of the septa are indicated in the 

 figure by the thin marginal lines and the oval interseptal 

 loculi. The two median septa are now distinct from one 

 another and the upper is larger than the lower. Duerden 

 indicates that the former is 



" thereby already recognizable as the main or chief septum." 



Figure 2. L. proliferum. Sec- Figure 3. L. proliferum. Section 



tion through the tip of a second of a third corallum from a still higher 



«orallum ; a little higher than in level. After Duerden. 

 figure 1. After Duerden. 



He also makes the important observation that the 



" upper primary interseptal spaces are much narrower than the 

 middle and lower interspaces. Within each of the latter an 

 additional septum (A) has appeared, and within the middle, right 

 interspace the rudiment of a second additional septum (B) occurs. 

 No new septa are ever developed within the two upper intersep- 

 tal spaces." 



Figures 4 and 5 are easily interpreted with respect to what 

 has already been explained concerning the structural features 

 of previous figures. Figures 6 and 7 will be referred to again. 



The reader's attention is now directed to figures 8, 9, 10 and 

 11, which are Duerden's diagrams reproduced again, but this 

 time inverted. This has been done for the sake of easier com- 

 parison. In figure 8 it will be noticed that what are now the 

 upper lateral septa are, as indicated by the dark lines, more 

 nearly at right angles with the median than are the lower ones, 

 while the median dark lines of the lower ones in this figure 

 (upper in Duerden's) have a proximity to the median dark line 

 which is at this stage very noticeable and suggestive. 



Before entering upon a further discussion of the diagrams 

 obtained by inverting Duerden's figures, I wish to call atten- 



