322 Variatiun and C'orrelation in Arcella 



are of course very grcatly reilucod, and consequently it bccomes possible to 

 measure a large series uf individuals. 



Besides the deterniination of tho coiistanls of the simple Variation iu single 

 dimeusions \ve also wishcd to ascertain the auiount of the corrclation in the 

 Variation of different parts. As iu the case of simple Variation Simpson's is the 

 only work which \vc have on correlation in any protozoan. His data enablcd him 

 to detormine, in addition to the correlation between the two individuals of the 

 divided pair, the correlation between length and breadth of the same individual. 

 His value for the length-breadth corrclation in the same individual was "421. It 

 is very iniportant to duterniino wliether there exists in general a higher, or lower, 

 or equal amount of correlation between the parts of lower organisms as compared 

 with higher. This question of the relative amount of correlation in different 

 fornis bccomes very significant wheu any attenipt is madc to solve the more 

 general problem of the physiological basis of the phenomenon of correlation in 

 Variation*. Although the fundamental charactcr of this general problem is 

 apparent, yot comparatively little has been doue towards solving it. It seems not 

 unlikely that the ultimate method of attacking this problem will be by means of 

 experimcnt, but some light can certainly be thrown on the question by deterraining 

 the amount of correlation in Variation in different groups of organisms. Such 

 determinations can then be used as a definite quantitative basis, from which to 

 proceed by experimental methods. It has seemed to one of the writers that one 

 good starting point for the investigation of this problem would be a thorough 

 study of Variation and correlation in the cell, including both free-living and tissue 

 cells. The present paper forins uur first contribution to this study. 



II. Material and Methods. 



The form u.sed in this work was the common Arcella, A. vulgans. A very 

 complete description of this organism, illustrated with coloured figures, is given by 

 Leidyf . It is also fully described by Bütschli:]:. 



Arcella is so generally distributed am! well known an organism as to make 

 only a very biief account of its taxonomie position, structure, etc., necessary here. 

 Arcella is a shelled rhizopod belonging to the group I mperfvrata, family Arcellina, 

 as defined by Bütschli. The shell is composed of chitinous material, and varies in 

 colour from almost purely colourless transpareucy to a deep brown. In surface 

 view the shell is seen to be composed of very small hexagonal Clements packed 

 closely together. The normal Arcella shell is cii-cuiar in outline, with an 



* For an historical account of the discussioii of the phenomenon of correlntion by carlicr workcrs 

 cf. Ha<il, E., '• Ueber die Bedeutung des Prinzips von der Korrelation in der Biologie." Biol. Centralbl., 

 Bd. XXI. pp. -lOl-lKi, i'Ji-VJl, 5.50-560, 585-591, 605-021, l'JOl. 



t Leidy, J., " Frcshwater Ithizopods of North America." Jiep. U. S. Geol. Surv. of the Territoriet 

 (Ilaydeu). Vol. xii. 1879, pp. 166-173. I'late XXVII. 



t Bütschli, 0., "Protozoa," iu Bronn's Klassen und Ordnungen des Thierreiches. Bd. i. 1 Abth. 

 188Ü-82. 



