390 TRANSACTIONS AND PROCEEDINGS OF [Sess. lvi. 



contains a coiled filament, which, at the commencement of 

 division, breaks up into six {Helix) or eight {Limax) rods ; 

 that these in their turn divide longitudinally, move asunder, 

 and form the chief rays of the polar radiation. Prenent * 

 saw also ribbon-shaped filaments in the spheres attractifs, and 

 considers them as rudimentary forms of the paranuclei. 

 Hermann t has seen twelve crescentic filaments in Helix, and 

 also groups of short S-shaped or looped filaments surrounding 

 the " central body " in the spermatocytes of Proteus anguineus. 



Hermann has further, t in his description of the sperma- 

 togenesis of the Salainander, pointed out that in the para- 

 nucleus a globular body (the middle piece of the ripe 

 spermatozoon) is stained red by saffranin, while a ring-like 

 body (the spiral membrane on the tail of the mature sperma- 

 tozoon) is stained violet by gentian-violet. In the same 

 paper a method of differential nucleolar staining is given, 

 namely, the nucleoli red, by saffranin, and the rest of the 

 cell violet, by gentian- violet, in resting cells, while during 

 division monasters and diasters are stained red, and every- 

 thing else stained violet.§ The author, besides mentioning 

 the facts of the differential nucleolar staining, and that a 

 part of the paranucleus is stained violet while another part 

 is stained red, has either not deemed it prudent to draw a 

 comparison between the staining reactions of the nuclei and 

 paranuclei, or has overlooked the fact. 



I believe the fact just stated, that it is possible to stain 

 the paranucleus analogously to the nucleus, to be of the 



* La Ctaiule, IV. 1. 



f Archiv. f. Mikrose. Anat. 37, p. 585. 



X Hermann, Hodenstiiilien' in Aicliiv. Mikrosc. Anat. xxxiv. 



§ Gnignavd was tlic lirst to put on record a method for difi'crontial niicleolar 

 staining with metliyl-grecn and i'uclisin, or with ha?matoxylin and saffranin, 

 in 1885 [Ann. Sc. Nat. ser G, T. xx. \). 318]. Hermann comes next with liis 

 sad'ranin and gentian-violet method, in A]>ril 1889 [Arcliiv. Mikr. Anat. 

 xxxiv. ]). 60]. I then ])nhlished an account in Jan. 1891, Trans. I'ot. Soc. 

 Edinb. , showing that dillercntial stnins miglit he ^nodiieed Ijy lieliocin and 

 nietliylene-hlue, or eosin (crythrosin) and iiKilliyleiie-bhic, or Jiigrosin and 

 eosin, or nigrosin and hematoxylin. 



Flemming, in his latest paper, Nenc Bt'itnige z. Kennt. d. Zelle, 24th April 

 1891, Arch. Mikrosc. Anat. xxxvii. p. C97, says that he demor.strated 

 double stains as far back as 1884, at the Kopenhagen Medical Congress, and 

 that since that time he lias fre(]nently demonstrated tlie same phenomenon. 

 Flemming uses for differential staining, saffranin, ha-mato-xylin, saffranin- 

 mannein, or .saffranin-gentian-violet. 



It is evident that I knew of Guignard's method, but not of Iliumann's ; 

 that Hermann nnd Flemming also, quite inde])endently of one another, made 

 the same observations as to the different behaviour of the nucleolar and nuclear 

 matters in regard to staining reagents. 



