597 



be accepted unless it can be shown that bodies of the shape of the 

 corpuscles, formed of a uniform mixture of nucleoproteids, lecithin and 

 Cholesterin in the proportions in which they occur in the envelope, 

 have the same tendency to adhere into rouleaux in Norris' experiment, 

 as when the discs are covered with a film of fatty material. 



The envelope is therefore a composite one being probably formed 

 internally of nucleoproteid matter, externally of lecithin and Cholesterin 

 (myelin). It is affected by all agents which atfect either of its com- 

 ponents. Such agents, even if they dissolve only one of these com- 

 ponents, may influence the porosity of the membrane, and permit the 

 escape of the coloured fluid contents, without necessarily producing 

 a rupture of the membrane. Other agents such as alkalies, which 

 dissolve both components of the envelope, produce an instant dis- 

 appearance of the whole corpuscle. 



I do not propose in this place to deal with the special fibrillar 

 structures which have been described by Meves ^) at the circumference 

 of the amphibian erythrocyte. That these are true fibrils and not 

 merely an appearance due to folds in the membrane as has been 

 supposed by Weidenreich is proved by the observations of Bryce^), 

 who has exhibited them in sections across the erythrocytes of the 

 embryo Lepidosiren, where they lie within a clear space which oc- 

 cupies the rounded angle formed by the membrane at the edge of 

 the erythrocyte, and appear in transverse section as fine stained dots. 

 Bryce's figures and photographs show that at the stage of development 

 of the embryos investigated by him, the erythrocytes, although to all 

 appearance, when examined in the fresh state, of the same structure 

 as in the adult animal, nevertheless in fixed specimens show a con- 

 siderable amount of protoplasm extending in the form of anastomosing 

 strands between the membrane and the nucleus, and their richness in 

 protoplasm is further indicated by the fact that, like the red disks of 

 the Salamandralarva, as long ago described by Flemming, they are 

 capable of multiplication by karyokinetic division, which is not the 

 case with the erythrocytes of the adult animal. That such strands, if 

 present at all, must be very tender and easily ruptured in the corpuscles 

 of the adult amphibian is evidenced by the fact that the nucleus be- 

 comes easily displaced from its central position and may often be seen 

 lying excentrically in corpuscles of the blood of the newt (Triton cri- 

 status) when examined fresh and without addition of reagents: this 

 observation again affords the strongest possible testimony in favour of 



1) Anat. Anzeiger, Bd. 23, 1903, p. 212, and Bd. 24, 1904, p. 468. 



2) Trans. R. Soc. Edinburgh, Vol. 11, 1904, p. 291. 



