ZOOLOGY — LILLIE. 28 1 



netized needles. Groups of these when brought close together by 

 the action of a centrally attracting magnetic force are found to exhibit 

 arrangements closely similar to those shown by chromatic filaments 

 and chromosomes in the dividing cell. In this manner different types 

 of spireme formations, equatorial plates (both those with chromo- 

 somes arranged in a ring and those with chromosomes uniformly 

 distributed over the area of the plate) can be imitated. The aggre- 

 gation of the chromosomes in a single plane midway between the two 

 polar areas can be simulated by the combination of a central attract- 

 ive and two polar repellent magnetic poles. The inference is that the 

 chromosomes exert electrostatic repulsions on one other (due to their 

 being colloid aggregates of like sign) and that this mutual repulsion 

 is the chief factor in determining their disposition ; also that the astral 

 centers in mitosis repel the chromosomes. Since the chromosomes 

 carry negative charges, the astral centers must represent electrically 

 negative areas. There is therefore an electrical field within the cell 

 to which the disposition of the astral rays, etc., is regarded as due. 



II. Action of Salt Solutions on Ciliated Epithelium. 



The gill filaments of Mytilus edulis were used in this investigation. 

 The chief general results are as follows : 



( 1 ) Of alkali salts in pure solutions of single salts {e. g. , chlorides) 

 the order of destructiveness is as follows : Li, Na, NH^, K. In 

 solutions of K salts activity may last several hours. Cs and Rb 

 resemble K in action. 



(2) Na salts are peculiar in that their destructive action is readily 

 counteracted to a greater or less degree by the addition of various 

 salts with bi-, tri-, and tetra-valent cations. Anions are inactive. 



(3) Trivalent and tetravalent salts exercise this antitoxic action 

 in much lower dilution than bivalent, and tetravalent in somewhat 

 lower dilution than trivalent. 



(4) In the case of salts with bivalent cations a relation apparently 

 exists between the decomposition voltage of the cation and its poi- 

 sonous or antitoxic action. The valence of the cation seems to be a 

 factor of equal if not of greater importance in this action. 



(5) The toxicity of Na salts with different anions is not antag- 

 onized with equal readiness in all cases. In general those salts whose 

 anions have high decomposition- voltages are least toxic in pure solu- 

 tions, and their toxicity is antagonized most readily. The posses- 

 sion of a low decomposition-voltage confers a corresponding toxicity 

 on the anion and the salt can not be so readily antagonized. 



(6) Salts of certain bivalent metals have a destructive action which 

 is antagonized by the addition of other salts with powerfully acting 



