DICELLOGRAPTUS. 137 



designation of the axillary angle, and the space (axillary space) included between 

 the dorsal walls of the stipes is denominated the axil. The form of the axil, 

 which depends upon the direction and amount of growth of the proximal thecas, 

 may be rounded (J). Morrisi), square (D. complanatus), or angular (P. sextans). 



The sicula is of the same type as that of Leptograptus. Its apex (axillary 

 spine of some authors) is generally visible, occupying the centre of the axil. It can 

 occasionally be seen passing into the thread-like nema ; more frequently, however, 

 the apex is broken off, and then the sicula appears as a blunt node. The apertural 

 spine (virgella) of the sicula is usually conspicuous, occupying the central (initial) 

 position on the ventral margin, and is sometimes referred to as the " initial spine " 

 (the so-called radicle or radicular spine of earlier authors). 



The development of the proximal end of the polypary is essentially similar to 

 that in Leptograptus as regards (1) the presence of two crossing canals, and (2) the 

 alternate development of thecas l 1 , 1", 2 1 and 2 2 . In all the Leptograpti these four 

 earlier thecas all grow in a horizontal direction. In the widely divergent species 

 of Dicellograptus, such as I). Forchammeri, a similar mode of growth obtains; 

 but in the narrowly divergent (or more or less convergent) forms, such as 

 IK anceps, the distal portions of thecal 1 and 2 3 assume an upward direction 

 of growth, so that the form of the proximal end corresponds with that of 

 Dicranograptus. 



Throughout the entire genus thecas l 1 and l 2 are generally small; they each 

 make a sharp upward bend at a certain distance below their apertures, and at 

 the point of bending a single spine is given off. The two spines belonging to 

 these two earliest thecas are always conspicuous on the proximal end of the polypary, 

 and owing to their position with regard to the central or initial spine (virgella), 

 were termed " lateral " spines by the earlier authors (this title may be conve- 

 niently retained in diagnoses). In certain states of preservation these lateral 

 spines appear as if they were true apertural spines, such as are found in 

 Leptograptus, but in better preserved specimens it appears tolerably certain that 

 this appearance is deceptive, and that they are ventral and mesial in position. 

 These "lateral" spines are occasionally so strongly developed (l>., var. omatus) as 

 to give ground for suspicion that the two primal thecas, in their later stages, at all 

 events, became functionless. 



In many species only a few of the more proximally situated thecae exhibit 

 mesial spines similar to the lateral spines of thecas l 1 and l 2 , but their apparent 

 absence on the more distal thecas may be merely a result of their small size or 

 imperfect state of preservation. In one species — D. anceps — all, or nearly all, the 

 thecas are spinose, but this is only evident in exceptionally well preserved 

 specimens. These thecal spines are never apertural in origin, but, like the lateral 

 spines of thecas l 1 and l 2 , develop from a point in the ventral wall of the fcheca. 

 In some species (D. Forchammeri) tliis point is immediately below the aperture; in 



