THE AMPHIBIA OF OKLAHOMA 



per unit of time between their means (Table III, items 6 and 7), but the 

 number of observations was inadequate to show much. Where two or more 

 items record the same water temperatures but different air temperatures (or 

 vice versa), one variable is controlled; if, therefore, the rate varies directly 

 with the other variable this should be evident. In items 2 and 9, WT =: 11.00 

 and the difference in AT ^^ 2.5°. The mean call rate is 2.1 higher at the 

 higher temperature. This corresponds to an average of 1.5-|- increase in call 

 rate per degree rise in AT. But with only one degree less in water tempera- 

 ture (items 2 and 9) the corresponding figure is only 0.1 — and other items 

 (e.g. 10, 13, 14) correlate poorly with this. If we arrange means according to 

 increasing air temperatures, disregarding temperature of water, we come out 

 with the following sequence: 10.6, 13.4, 9.9, 12.4, 18.3, (14.5 and 23.9), (14.6 

 and 16.2), 22.2, and 24.2. A similar sequence for water temperature disre- 

 garding that of air is: 13.4 (14.5 and 12.4), (18.3 and 10.6), (19.0, 14.5, and 

 16.2), (22.2 and 23.9), and (24.2 and 29.0). Neither of these sequences of 

 means nor similar sequences of modes indicates that rate of call is wholly a 

 result of either air or water temperature taken alone. If we assume that the 

 rate rises with increase in both air and water temperatures and that these 

 have equal effects, then in a sequence of means of air and water temperatures, 

 means in rate should also consistently increase. Such a sequence is 13. 4, 10.6, 

 12.4, 18.3, 14.5, 19.0 (14.6 and 16.2), 22.0 (24.2 and 29.0), and 23.9. This is 

 also negative. But if one take the means of each successive three of this 

 sequence, thus further smoothing out individual variations, we do come out 

 with an increasing rate as follows: 12.1, 17.3 — , 17.3, 25.4 — . It is very ques- 

 tionable whether 252 cases (omitting items 1 and 1 1 in the table) are sufficient 

 data to justify conclusions based upon this result. 



We therefore conclude that prevailing temperatures (within certain 

 ranges, at least) do influence the call rate of Ps. strecl^eri, that this influ- 

 ence is probably differential as between water and air, that other factors oper- 

 ating in individual cases in congresses modify these effects, and that many 

 more data are needed before a more adequate analysis can be made. These are 

 substantially the conclusions reached earlier (Bragg, 1942) where social fac- 

 tors were emphasized as involved materially. 



As stated above, the duration of calls and interval between calls in Ps. 

 styccl{eri are so short that I have been unable to record them accurately by 

 the method used. To my ear they usually seem to be of about equal length 

 for any one individual, varying of course with the rate of calling at any one 

 time and place. At lower temperatures, however, I have sometimes suspected 

 that the rate of call was lower than usual because of a slightly lengthened 

 interval, the duration remaining nearly constant. I hope to secure data to 

 test this matter in the future. 



The data tor Ps. triseruita essentially substantiate conclusions drawn in 

 an earlier paper (Bragg, 1948) and hence are not discussed in detail. It is 

 well to emphasize again, however, the marked difference in the calls of Ps. 

 cUirl{ii aiul Ps. triseruita. 



