We present results from a two week multi-longitude photometric campaign on TV Col held in 2001 January. The data confirm the presence of a permanent positive superhump found in re-examination of extensive archive photometric data of TV Col. The 6.3-h period is 15 per cent longer than the orbital period and obeys the well known relation between superhump period excess and binary period. At 5.5-h, TV Col has an orbital period longer than any known superhumping cataclysmic variable and, therefore, a mass ratio which might be outside the range at which superhumps can occur according to the current theory. We suggest several solutions for this problem.

We report precise measures of the orbital and superhump period in 20 more dwarf novae. For 10 stars, we report new and confirmed spectroscopic periods-signifying the orbital period P-o-as well as the superhump period P-sh. These are GX Cas, HO Del, HS Vir, BC UMa, RZ Leo, KV Dra, KS UMa, TU Crt, QW Ser, and RZ Sge. For the remaining 10, we report a medley of P-o and P-sh measurements from photometry; most are new, with some confirmations of previous values. These are KV And, LL And, WX Cet, MM Hya, AO Oct, V2051 Oph, NY Ser, KK Tel, HV Vir, and RX J1155.4-5641.Periods, as usual, can be measured to high accuracy, and these are of special interest since they carry dynamical information about the binary. We still have not quite learned how to read the music, but a few things are clear. The fractional superhump excess epsilon[= (P-sh - P-o)/P-o] varies smoothly with P-o. The scatter of the points about that smooth curve is quite low, and can be used to limit the intrinsic scatter in M-1, the white dwarf mass, and the mass-radius relation of the secondary. The dispersion in M-1 does not exceed 24%, and the secondary-star radii scatter by no more than 11% from a fixed mass-radius relation. For the well-behaved part of epsilon(P-o) space, we estimate from superhump theory that the secondaries are 18% +/- 6% larger than theoretical ZAMS stars. This affects some other testable predictions about the secondaries: at a fixed P-o, it suggests that the secondaries are (compared with ZAMS predictions) 40% +/- 14% less massive, 12% +/- 4% smaller, 19% +/- 6% cooler, and less luminous by a factor of 2.5(7). The presence of a well-defined mass-radius relation, reflected in a well-defined relation, strongly limits effects of nuclear evolution in the secondaries.

We report photometry of the helium-dominated cataclysmic variable HP Librae during 1995-2001. The main photometric signal varies between 1118.89 and 1119.14 s, on a timescale of a few years, and displays a waveform characteristic of "superhumps." After subtracting the main signal, we found a weak residual signal at s, which we interpret as the underlying orbital period of the binary. The full amplitude of this 1102.70+/-0.05 putative orbital variation is just 0.005 mag, the weakest orbital signal yet found in a cataclysmic variable (CV). The 1119 s signal of HP Lib is a superb match to the well-studied 1051 s superhump of AM CVn, the "mother of all helium CVs." The superhump shows no change in amplitude or waveform on any timescale and no essential change in period on timescales shorter than similar to3000 cycles. Such great stability makes the star a promising test case for detailed studies of the underlying spiral structure in the disk, the likely cause of superhumps. Comparison of orbital and superhump periods for the family of AM CVn stars supplies evidence that these stars are evolving toward longer orbital period.