378 NATURAL SCIENCE. Dec. 



is given off each tetrad divides into an extruded and a retained dyad. 

 For the second polar body, each of the retained dyads divides, and 

 three chromosomes are extruded, three retained, each being a quarter 

 of the original tetrad. It is clear that the distinction between 

 reducing divisions and ordinary divisions suggested by Professor 

 Weismann cannot be applied to the formation of the polar bodies as 

 interpreted by Professor Wilson. Another interesting general point, 

 very plainly illustrated in these photo-micrographs, refers to the early 

 stages in polyspermy. In Toxopneustes two spermatozoa may enter the 

 e^g simultaneously. An aster is formed around each, and they move 

 equally towards the nucleus of the egg-cell. When they reach it, 

 each aster divides, and there is formed a " quadrille " figure, consisting 

 of the egg-nucleus found with the two sperm-nuclei, and surrounded 

 by four asters at equal distances. The joint nucleus then divides, so 

 that four daughter-nuclei are formed. Professor Wilson thinks, and 

 his figures bear him out, that when Fol originally described the 

 " quadrille " figure, he was dealing with an abnormal occurrence of 

 this kind. 



Fossil Cells. 



It is plain that, though the hard cell-walls of extinct plants may 

 leave evidence behind them, as indeed they do in abundance, it is very 

 unlikely that there should be fossil evidence of animal cells. How- 

 ever, in some cases, it happens that such evidence does exist, and as 

 it may not be known to many of our readers, we give it. The earliest 

 known remains of the skeleton of chordate animals, which may have 

 been either fishes or something lower, are fragments of dermal 

 armour from the Silurian Formation. Their microscopical structure 

 is often beautifully preserved, and can be examined almost as readily 

 as the tissue of an existing animal. The simplest of these fossils are 

 thus proved to be mere granules of " shagreen," each formed round a 

 little papilla and nourished by radiating, branching capillar}^ tubules. 

 The soft papilla is never fossilised, always replaced by rocky matrix ; 

 but, judging by analogy with modern sharks, it must have been 

 furnished with the little formative cells termed " odontoblasts." 

 Sometimes these tubercles are fused together into continuous plates ; 

 and among Devonian fossils instances are already common in which 

 a great basal layer of hard tissue is developed, leaving the original 

 tubercles as a simple pustulate ornament on the exterior. The 

 secondary basal layer is often provided with little spaces, which must 

 have been occupied by cells similar to those of modern bone ; but in 

 very many — perhaps in the majority of cases — the tissue consists of 

 parallel or concentric layers conforming with the outer surface, and 

 not traversed by any passages analogous to the haversian canals of 

 true bone. Such skeletal tissue is named " isopedin." Typical bone, 

 however, is also found in some dermal armour even of Devonian age. 



Another form of dermal skeleton of the Upper Silurian and 



