is taken from V. Barger and D. Cline, Phys. Rev. Letters 16, 913 (1966); and to be published. The first four resonances are chosen for AL and A^ f and for N a the AT(938), iV*(1688), and a possible f ~ resonance mentioned in Ref. 8 are taken. 8 C. B. Chiu and J. D. Stack, Phys. Rev. 153, 1575 (1967). Our results reproduce their W = 0 intercept of the N a trajectory. 9 The N a and ATy trajectories do not turn around, while AQ turns around at W « 35 BeV and a « 125. Whether or not the trajectories turn around depends on the detailed asymptotic behavior of Ima^. Our results for the polarization in question (where W £ 5 BeV) are insensitive to such details o 10 Since Imc^ is small compared with Rea^, the imaginary part of Xi is ignored. n C. B. Chiu, R. J. N. Phillips, and W. Rarita, Phys. Rev 0 153., 1485 (1967).A binding energies of hypernuclei in the p shell follow a linear trend when plotted against baryon number ,4 ( Fig. 1). The over-all linear increase with A is attributed, in various approaches, 1 ? 2 to the strong spin-independent central part of the AN interaction. For each p -shell nucleon, an average of about 1-to 1.5-MeV binding has been obtained. 2 ? 3 The spindependent components of the AN interaction are expected to produce deviations of about 0.5 MeV 3 > 4 from the straight line shown in Fig. 1.There are some exceptions to this general description of p -shell hypernuclei, notably at A numbers 9 and 13. Both A Be9 and A c * 3 lie considerably below the straight line of Fig. 1. For these two A values surprisingly large binding-energy differences are found 5 " 7 :AB ( A Li 9 -Be 9 ) = 1.62 ± 0.19 MeV, AAA '(1) AAA It is to be noticed that # A ( A C 13 )= 10.9 ±0.3 MeV 7 is even lower thanB A ( A B 12 )= 11.06±0.14 MeV, 5 in spite of the smaller A value of A B 12 . The third example we wish to consider is that of 12 One can use more complicated expressions for X${W) which would, among other things, take account of its threshold properties. However, experimental information on these quantities at low energies is not very good and, consequently, no useful purpose is served in increasing the number of parameters. We have implicitly assumed that, asymptotically, Xf falls off as an exponential. This need not be true in practice.13 This is partly due to the fact that higher partialwave amplitudes carry larger statistical weight.14 The direct-channel contribution is comparable with p only around the dip region, where the p amplitude is ~J-its forward value. However, p is a rather weak secondary trajectory because at £ = 0 it is already ~^j the Pomeranchuk (P) amplitude. Therefore, compared with P the direct channel is quite small and the basic Regge hypothesis of the asymptotic dominance of the leading poles in unhampered.the A = 8 hypernuclei. The "exotic" A He 88 lies considerably above the straight line of Fig. 1. Here we have 5 ? 8 AS (He 8 -Li 8 ) =1.0 ±0.8 MeV,(2) AAA ' which might appear less problematic than in the previous examples (1). However, by coupling the particularly low first excited stat...