350 MUSCARDINID^E 



other respects their habits appear to be somewhat similar to 

 those of Muscardinus, as described below. 



The tail is very brittle in some forms, as in Elio?nys, 

 and breaks off, if seized, like that of a lizard, to the great 

 advantage of the escaping animal (see Helm, Zool. Garten, 

 July 1887, 217-219; Zoologist, 1888, 14-16). Thomas has 

 shown (Proc. Zool. Soc, London, 1905, ii., 491-494) that these 

 dormice further resemble lizards in being able to regenerate 

 their broken tails, and suggests that this power may be common 

 to "all species," but it does not yet seem to have been 

 demonstrated for Muscardinus. 



The dormice have been associated with the squirrels on 

 account of their climbing habits, and as often with the mice 

 and rats on account of their general external resemblance and 

 certain peculiarities of structure. But they have no close 

 affinities with either family, and stand naturally by themselves 

 as an intermediate group (see Forsyth Major, Geol. Mag., 

 November 1899, 49 2_ 495)> being absolutely differentiated from 

 both squirrels and mice by the number and structure of their 

 molariform teeth, and from all other rodents in the absence of 

 a caecum to the large intestine. Their murine appearance may 

 evidently be due to convergence, and is not necessarily 

 indicative of relationship. Although climbers, they are not by 

 that fact brought any nearer to the squirrels, and there are 

 many arboreal forms amongst the murines. Their feet also 

 are distinct from those of the squirrels, in which the claws are 

 more prominent, whereas in the dormice the small claws but 

 long digits with swollen pads enable the whole hand to be used 

 as a grasping organ, and also break the shock of a jump. 



These thickly padded feet are, however, strongly suggestive 

 of the harvest mice {Micromys), as are also, to a certain extent, 

 the ears, the valves of which are more complicated in structure 

 than those of typical mice or of squirrels. But here, again, 

 convergence, and not any true relationship, may be supposed, 

 the climbing habits of Micromys and of Muscardimis being 

 somewhat similar. 



Dormice fill a quite different role from that of the squirrels, 

 usually restricting themselves to creeping about in bushes and 

 undergrowth, and not ascending high trees. They are also 



