V\. PkARL and f. J. DlNBAR 32") 



probleiii of roeoriliiig defiiiitely flic culour ufsucli :ui orgniiisin ;is Arcella is ;i very 

 diffifiilt (iiic. Aft-LT ti-yiiig' and abaiuloiiiiig scvcral iiietliods of recording colour wo 

 finally dccided iipou the tüllowing as sufüicienlly accnrato for our pui'pose. In tho 

 work of Leidy alrcady referrcd lo, Plate XXVII contains a inimber of figures of 

 colour varieties oi Arcella. These figures are exlreniely well chosen to ty^pify the 

 different colour classes, ranging all the way fioni colourless transparcncy to a very 

 deep brown. For our colour determinations we chose four of these figures on this 

 Plate as types with whieh to conipare the colour of the Arcellae observed. As 

 typical of the lightest shells we took Fig. 14, and called this our colour type A. 

 The colour of this figure on the Plate is a very light yellow-browu. It very well 

 represents the colour of the lightest shells wo found. For the next darker type we 

 chose Fig. 7, and designated this as colour type B. Shells falling within this type 

 have a distinct yellow-brown colour, and yet are still fairly transparent. Fig. 19 

 served as our colour tyj)e G. Shells falling in this colour class may be roughly 

 described as of mediuni-brown colour, with a sensible though sniall admixture of 

 yellow. As representative of the darkest individuals, our colour type D, we chose 

 Leidy 's Fig. 2. This is the colour inost frequently observed in Arcella shells, and 

 is a dark, even brown. 



As the Arcella shells were measurcd their colour was compared with these 

 figures and oach shell was recorded as .4, 5, G or D, according to the figure which 

 it most noarly niatched. It was found in jjractice that the colour of the shells could 

 be more accurately classified in this way than by coniparing with sheets ofcoloured 

 paper for exaniple, because the thing with which the colour was compared in this 

 method was an actual representation of an Arcella. While the method gives us 

 no quautitative determination of the amount of pigmentation it does furnish a 

 reasonably good Classification of Arcella shells by colour, which can be used for 

 purposes of correlatiou. This was all we hoped or desired to get in the present work. 



The 504 individuals of which measurements were niade were all taken from a 

 Single small laboratory culture dish. The material was collected from a srnall lake 

 in the vicinity of Ann Arbor, and the few Arcellae originally in the collection 

 multiplied rapidly under the conditions of the laboratory culture*. At the time 

 when the measurements were begun the Arcellae were very abundant. Since the 

 individuals were all taken from one sniall culture dish they may be considered a 

 very homogeneous sample of this Arcellae population, as far as environmental 

 conditions are concerned. The sample is not, however, quite homogeneous with 

 respect to age, for the following reasons. Pressure of other work made it necessary 

 for US to extend the time of making measurements over a period of about three 

 weeks. It is likely that some of the last individuals to be measured represented 

 later generations than those to which those measured earlier belonged. That this 

 was the case is indicated by the form of the frequency polygous obtained. It will 

 be shown farther on in the paper, however, that there was on the whole no sen.sible 

 increase in size of the shells duriiig the time which the making of the measure- 

 ments covered. 



* [Since the original Arcellae were few in number, it sliould be notcd tlmt the authors are dealing 

 ratber with a few large " fraternities '' thau a geueral population of A icellue ; such, however, is probably 

 the Constitution of most " cultures." Ed.] 



