432 STUDIES IN EVOLUTION 



it a twelve-sided prism. This stage is the third toward the 

 formation of a series of mature interstitial calices. 



Durinof the third stas^e the intermural buds increase in 

 size until they completely surround tlie parent cell. Then 

 further growth truncates their adjacent angles, thus adding 

 two more sides to each bud, making them pentiigonal in 

 section. This marks the fourth stage of intermural growth. 

 At the same time the central corallite loses six of its sides, 

 and returns to its early hexagonal form. The axes have 

 revolved 30°, and the original sides have now become the 

 angles of the corallite (Plate XXXII, figure 4). 



At this period of growth it is necessary to consider a 

 series of buds on the periphery of the corallum, marked i", 

 2'\ etc., in Plate XXXII, figures 3 and 4. They are first 

 triangular in form like the others, and of two sizes, owing to 

 their different ages. The growth of this series continues 

 until they touch and truncate the angles of the first series 

 (i', 2\ etc.), producing the fifth condition or stage. The 

 first series of buds has now three hexagonal and three 

 pentagonal corallites (Plate XXXII, figure 5). 



In the last or sixth stage (figure 6), the further growth of 

 all the intermural cells results in a corallum of nineteen nearly 

 equal hexagonal corallites. The original parent cell (A) is 

 at the centre, the first six intermural cells (i', 2\ etc.) com- 

 pletely surround it, and the six new peripheral corallites 

 (i", 2'\ etc.) are interposed between the members of the 

 original circlet (i, 2, etc.). The effect of this intermural 

 growth, then, is to dissociate all the first series of corallites 

 from the parent cell and from each other. (See Plate 

 XXXIII.) 



The changes taking place in the number and form of the 

 cells may be tabulated as on page 433. 



Buds are developed in Favosifes and Michelinia whenever 

 there is a space or opportunity for their growth, unless the 

 corallum is affected by some abnormal condition. If this ten- 

 dency to form a solid mass of corallites were not so strong, 

 and if tlie process of budding took place only at compara- 



