Accounting for endothermy in fishes.

of a male. The number of mounts attempted by a male was also scored and recorded. The results of all the experiments were converted to lordosis quotient (LQ), defined as a percentage of full lordosis response (perineum elevated, all four legs extended from the initial crouch position and the head at an angle of 450 from the floor) divided by the number of mounts. Each group consisted of six animals, and each female was tested for 10 min in the presence of a male. The animals were tested for lordosis response at 30 min and at 3 hours after injection of compounds. The effects of the compounds were not transient, because identical results were obtained at 30 min and at 3 hours after infusion. 29. Statistical analysis was done by either of two methods as appropriate: Kruskal-Wallis one-way ANOVA on ranks followed by Dunn's method for comparision of all groups versus control group, or one-way ANOVA followed by Dunnett's method for comparision of all groups versus control group. Sigma Stat (Jandel Corporation) was used. 30. We thank K. Mayo for the sequence of the rat PR mRNA, J. Turcotte for assistance with the binding assays, the staff of the Center for Comparative Medicine at Baylor College of Medicine, and D. Scarff for help in the preparation of the figures. Supported in part by N.S. 19327 (NIH) and Research Scientist Development Award grant MH-00885 to J.D.B. and a NIH (National Institute of Child Health and Human Development) grant to B.W.O.