TRENTON LIMESTONE. 65 



of Favositcs Lycoperdon.* Some of tin; hcniispheiical masses sliow a great divergence of the 

 tubes, as if radiating from an axis ; while others are nearly vertical. Specimens approaching 

 a conical or ramose form exhibit a tendency to radiation of the tubes from a central axis. 



A comparison of all tliese forms shows no essential difiercnce of structure, farther than 

 might be expected from the diffcront conditions under which the fossil was developed ; and 

 I am al present disposed to regard all the forms referred to this species as varieties of the 

 same, although there may be some links wanting to show the perfect gradation from the 

 hemispheric to the ramose forms. It is not intended to be understood that the hemispheric 

 and globular masses always develop the ramose varieties, like fig. 2, pi. 23 ; these doubt- 

 less having this form from their commencement. Still the specimens 1 g, pi. 23, and 1, 

 pi. 24, show that the massive varieties do sometimes become ramose to some extent ; and 

 there are other specimens, which could not l)c represented in figures, that prove more 

 clearly the development of the ramose, from the massive forms. In addition to these more 

 regular forms, this coral sometimes presents itself in irregular massive forms, as if, while 

 a group had been growing, they were distmbed, a part of the polyps being destroyed, and 

 the others growing onward and increasing, finally producing shapeless masses. 



These remarks are made for the purpose of calling the attention of the student to these 

 circumstances, particularly as the hemispheric masses alone have been referred to Say's 

 species. These observations will be better understood by a comparison with the following 

 figures, drawn from different specimens. 



Fig. 1. Portion of a vertical section of a hemispherical iorm of large size, showing the radiating 

 structure from increase of tubes by subdivision, and also the increase by development of 

 marginal tubes. The tubes are filled, and have become solid columns, which are easily 

 separable by a slight blow of the hammer. 



Fig. 1 a. A portion magnified, showing only the enlarged columns; the diaphragms not visible. 



Fig. 1 b. Transverse section of a small conical or hemispheric form, showing the radiating arrangement 

 of the tubes. 



Fig. 1 c. A portion of the same enlarged, showing the diaphragms at regular intervals in some of the 

 tubes which are divided longitudinally. 



Figs. 1 d, e, /. Different external forms of the coral : 1 d and / with more depressed forms, being the 

 more usual ; while 1 e shows a tendency to branching, which, if continued, would produce 

 a coral with a massive centre and numerous branches. 



Fig. \ g. A. large irregular mass with numerous diverging branches, some of which arc broken off, 

 showing the radiating arrangement of the tiibi s from a central axis. This specimen exhibits 

 the same tendency as in 1 c, carried forward to a greater extent, showing how the same 

 form may produce the ramose varieties. The tubes or solid columns, where the branches 

 are broken oflf", present precisely the same structure as in the smaller hemispheric masses. 



Fig. 1 A. A fragment from a larger mass, where the tubes are not filled with calcareous matter, showing 

 the regularity of the diaphragms, which are distant a little less than the width of the tubes. 



Fig. 1 i. The same enlarged. The mass is light and spongy like recent coral, and appears to be com- 

 posed of numerous smaller ones which were drifted together, and afterwards recommenced 

 their growth, forming a very irregular mass. 



* From its resemblance to Lycoperdon bovista. 

 [ PALiEONTOLOGV.] 9 



