358 J. K, BI,()MFIK!,D.- 



in shape, being sometimes pyiiform, at others caliciform, 

 but ill all the shrivelled remains of the nucleus can be 

 seen at the base. The cells of the most external row 

 sometimes present a beaded appearance on their free surface, 

 which reminds one of the external cells in Petromyzon, 

 which have their side much thickened. 



In the deeper parts of the unstained layer young goblet 

 cells may be seen, which possess a normally-sized nucleus, 

 and are capable of staining; and between these and the 

 mature mucus-cell all stages are recognisable — in fact, there 

 are few tissues in the animal kingdom that afford such good 

 examples of goblet cells and their formation (fig. 2 a). 



The ordinary epidermic cells need not detain us. The 

 lowest series (fig. 2 h) have an expanded base, and the higher 

 form transitions to the mucus cells ; but embedded among 

 them are two larger kinds, which merit a longer description. 

 The first kind are stained yellow by the picric acid (fig. 1 d). 

 They occur at all parts of the epidermis, and may sometimes 

 be seen protruding between the superficial cells at the surface. 

 In shape they are irregular, but always oblong, appearing 

 to have their form determined by the ordinary cells which 

 surround and firmly embed them, for it is impossible in teased 

 preparations to obtain them quite isolated. Their appear- 

 ance is coarsely granular, and no obvious nucleus is present, 

 but often a small red spot may be detected, which would' 

 appear to be the remains of that structure. From the fact 

 that they stain yellow with the picric acid of the picrocar- 

 mine one is lead to the idea that they correspond to the 

 " club cells " of Petromyzon described above ; and though 

 they are not so complicated in their structure, yet their 

 behaviour with picric acid and similarity of their fate— for, 

 like those cells, they evidently grow towards the surface and 

 are discharged-— leave little room to doubt that they are 

 homologous. 



The second kind of cell (fig. 2 e) alluded to above would 

 seem to correspond to the granular cells of Petromyzon, though 

 the correspondence is not so well made out. In most of my 

 sections, which were prepared from specimens hardened in 

 chromic acid or Miiller's fluid, these cells appear as clear 

 spherical spaces embedded among the ordinary epidermic 

 cells which are compressed and flattened in their immediate 

 neighbourhood. 



The periphery is bounded by a membrane, and in the 

 interior is a finely granular mass containing a deep-stained 

 spot, which appears to be the remains of a nucleus. From 

 the granular mass processes of similar substance radiate 



