32 FROPESSOIt LANKESTER. 



larity in rectilinear series, may be described as tbe rectilinear 

 tesselate form of aggregation. Sarcina, and more especially 

 Merismopedia, furnish very similar aggregates of their con- 

 stituent plastids. The gathering which I received from Mr. 

 Archer contained sheets of plastids in which the plastids 

 were of this hour-glass shape^ and arranged in this same rec- 

 tilinear manner; but the sheets were of larger size, consist- 

 ing of as many as twelve or twenty rows. The aggregates 

 of this kind very generally are nearly square plates, and have 

 so symmetrical an aspect as to suggest rather a Avork of art 

 than a natural growth. Frequently, at the sides of the square 

 plate, a group of four, six, or eight plastids has become de- 

 tached and fallen away, leaving a square hole or blank in 

 the series. I hesitated for some time about regarding these 

 regularly arranged plates as forms of B. rubescens, until the 

 study of Mr. Archer's specimens enabled me to satisfy myself 

 that the plastids are really coloured centrally with the cha- 

 racteristic purple-red pigment of that species, and also fur- 

 nished additional proof of the connection of this form with 

 the other varieties of B. rubescens, in the fact that they oc- 

 curred in his gathering (as they had in those I first observed) 

 in connection with a great abundance of other varieties assign- 

 able to that species, such, for instance, as fig. 16 and fig. 21 

 of Plate XXI II of my former memoir. 



The form of the individual plastids, though very strongly 

 marked, was already familiar to me. Some of this form are 

 drawn in figs. 13 and 26 of my former memoir. In fig. 6, 

 Plate III, 1 have represented a plastid of the same form, 

 which is interesting as giving a transition between this very 

 regular unilocular form and the multilocular forms. In ad- 

 dition to the central hour-glass-shaped cavity filled with 

 coloured material, a small eccentrically placed granule is 

 seen, eating its way, as it were, into the thick colourless 

 wall of the plastid. 



The plastids of the regular tesselate form of aggregation 

 cohere laterally by means, no doubt, of the gelatinous sub- 

 stance which forms a sheath to each of them. One would 

 suppose, from the analogy of tissues and cell- development 

 generally, that such a regular aggregate had been formed by 

 a regular process of cell division in situ. I am inclined, 

 however, to believe that this is not the case, and that the 

 tesselate aggregations of all kinds of Bacteria, as well as the 

 retiform (such as that drawn in fig. 19 of my former paper), 

 take their origin by the spontaneous apposition of indepen- 

 dently developed plastids, just as the network oiHydrodictyon 

 is developed. 



