REACTIONS OF FROGS TO CHLORIDES 611 



potassium is less stimulating than ammonium, though in the 

 weaker concentrations (1 m., m./2) the reverse is true. 



It is also evident in comparing the various averages in table 6 

 that individual differences in the groups of frogs must be taken 

 into account in interpreting the table. It is to be expected that 

 the reaction-time would shorten with increased concentration of 

 the stimulating solution and this expectation is realized in all 

 parts of the table except those referring to concentration 1 m. 

 Here the reaction-times are shorter than might have been an- 

 ticipated, but this condition is probably due, not primarily to the 

 solutions, but to the individual peculiarities of this particular set 

 of frogs. That frogs vary much in this respect, can be seen in 

 many of the tables; thus, of the four frogs whose reactions are 

 recorded in table 3, frog J is slowest in its average reaction-times 

 for all salts, and frog L is quickest. It is, therefore, very prob- 

 able that the unexpectedly short reaction-times at concentra- 

 tion 1 m. were due to the accidental use of four unusually rapid 

 frogs. This opinion is supported by the results (table 7) of test- 

 ing a frog with one salt at the four concentrations used. As will 

 be seen in the table the reaction-times lengthen with increased 

 dilution. 



In seeking for an explanation of the reaction-times called forth 

 by the different salts, one naturally turns to the ionic contents 

 of the solutions. The degrees of ionization at the concentra- 

 tion of the four salts used in these experiments are given in 

 table 8. 



According to the degrees of dissociation given in table 8, the 

 four salts experimented with fall into two groups, one consisting 

 of ammonium and potassium, and the other of sodium and lithium. 

 As I have already pointed out, the reaction-times (tables 1 to 5) 

 lead independently to a like grouping of these salts, so that these 

 times alone would enable the experimenter to distinguish these 

 two groups, especially at the concentrations 2 m. and 1 m. 



The relation, therefore, between the reaction-times and the 

 degree of dissociation is close, though as a comparison of tables 6 

 and 8 will show, it is not mathematically proportionate. Braeu- 

 ning ('04) has stated that in stimulation by salts the reaction-time 

 is not primarily a function of the diffusion process, as he believes 



