The tensor analyzing power T 2 Q for the radiative capture reaction 1 H (d po] ,y) 3 HQ has been measured in order to test new three-body wave functions. This observable arises from the D state of 3 He. An effective two-body direct-capture calculation, which was previously shown to fit the a 2 coefficient of a Legendre-polynomial expansion of the differential cross section, is found to give a good description of the present data. A value for the asymptotic D/S ratio is extracted from the data although it is found to be model dependent.PACS numbers: 21.40.-hd, 24.70. + s, 25.45.-z, 27.10. + h Recent theoretical calculations of few-body wave functions have stimulated interest in new measurements on very light nuclei. In particular several related reactions have been used to study the ground state of 3 He. Among these are photodisintegration 1 ' 2 and electrodisintegration 3 of 3 He and radiative capture of protons by deuterons 4 " 6 and of deuterons by protons. 7 The a 2 coefficient of a Legendre-polynomial expansion of the differential cross section for y-ray emission following capture of protons by deuterons has been found sensitive to the Z)-state component of the 3 He ground-state wave function. 5 As pointed out by Seyler and Weller, 8 the tensor analyzing powers of the reaction 1 H(rf pol ,y) 3 He are expected to be more sensitive to ZXstate effects. In fact they are identically zero if the S= y capture amplitude, which arises from the D state, is absent. Of the previous work only Ref. 6 involves a polarized beam, the incident particles being protons. Therefore, it is important that experiments be done with a polarized deuteron beam. Measurements of r 20 , with good statistical accuracy, have been performed for the reaction l H(d poh y) 3 He and are reported in this Letter. An effective two-body radiative-capture calculation, similar to the one used in Ref. 5 to describe the differential cross section, fits the present data well. In addition, an upper limit of 0.050 for the asymptotic D/S state ratio 17 is implied by the data.The deuteron beam used in this reaction produces a high neutron flux which makes it difficult to detect the y rays. To avoid this problem the recoiling 3 He nuclei were detected instead with an Enge split-pole spectrograph. A one-to-one relationship exists between the energy of the 3 He and the angle of the y ray (Fig. 1). However, for the beam energy used, the 3 He nuclei are confined to a cone about the beam axis with a maximum angle of 2.6°; the beam particles are therefore in the way. An appropriate magnetic field was chosen such that the beam was collected in a Faraday cup at a large radius of curvature R in the spectrograph, while the 3 He particles were separated magnetically and focused at low R. The McMaster University Lamb-shift polarized-ion source was used to produce polarized deuterons alternatively in the m = 1 and m = 0 substates relative to the beam direction. An energy of 19.8 MeV was used. The target was a polyethylene (CH 2 ) film approximately 100 /ng/cm 2 thick. The ...