398 Professor H. H. Swinnerton— 
units are not so immutably correlated with one another that the 
presence of one of necessity involves the presence of all the others ; 
thus the presence of the columella does not involve the presence 
of the tabelle, nor the dissepiments that of the septa. Had these 
units all been immutably correlated, then only one morphological 
series of corals would have been possible, and Carboniferous 
paleontology could then have been studied without tears. In actual 
fact the elements may be present or absent in varying combinations, 
and thus a wide variety of coral types may exist. The number 
possible is a matter for mathematical calculation. The real number 
may be smaller, for a coral consisting, for example, of tabule alone 
is inconceivable. 
THE MEASUREMENT OF STRUCTURAL UNITS IN CORALS. 
Not only do the structural units vary as to their presence or 
absence, but the individual elements themselves are also subject 
to perpetual change in quite definite directions. 
Fic. 5.—Portions of sections of mutations K, y, 5 from the gens Canina 
cylindrica (after Vaughan). 
A brief inspection of a number of coral-slices of several genera, 
or of different parts of the same individual, will show that the 
elements differ in the various slices. These differences may be sorted 
out and arranged in series. Thus the inner ends of the major septa 
(Fig. 5) may be withdrawn from or approximated towards the 
centre. Similarly the outer ends may extend to or be separated 
from the theca (Fig. 5). This incidentally furnishes an illustration 
of the same unit being affected by two divergent series of changes. 
In both cases the changes may be accurately recorded by measuring 
the distances of the inner and the outer ends of the septa from the 
epitheca. The figures may then be expressed as percentages of the 
average diameter of the coral. 
Where minor septa have become units of sufficient prominence 
they, too, may be measured in the same way. For both kinds of 
septa the total number present is a feature of some importance. 
The dissepiments occupy a wide or narrow zone which always lies 
in contact with the epitheca. The width of this zone is easily 
measured, and, again, may be expressed as a percentage of the 
coral diameter. 
