Measurements of the 16-Hz susceptibility of a powdered Na3lCe(C 7 H 3 N0 4 ) 3 ] • 15H 2 0 (CDP) sample, in the shape of a right-circular cylinder with diameter equal to height, have been made as a function of the Johnson noise temperature Tj of a resistor in the range 2 to 20 mK. Pure liquid 3 He was used for thermal contact between the CDP and the noise thermometer in a demagnetization celL The results are compared with similar data for cerium magnesium nitrate (CMN). CDP appears, like CMN, to have an anomalously low thermal resistance to pure liquid 3 He.In a recent Letter Doran, Erich, and Wolf 1 introduced the material Na 3 [Ce(C 7 H 3 N0 4 ) 3 ] t 15 H 2 0, abbreviated CDP, as a possible alternative to cerous magnesium nitrate (CMN) for millikelvin thermometry. CDP has both a smaller interaction specific-heat constant CT 2 /R and a smaller Curie constant than CMN and, according to Ref. 1, CDP reached a lower magnetic temperature on demagnetization than CMN. In order to utilize any superior properties of CDP, it is essential to make thermal contact to it. This is usually achieved with CMN by finely powdering it and mixing it with pure 3 He, in which case a magnetic coupling^3 is responsible for the low-temperature surface contact. CDP is less stable than CMN, and hence fine powdering might be a problem (no protective oils may be used when good thermal contact at low temperatures is desired). Further, it was not known whether the excellent thermal coupling possible with CMN is possible with CDP as well. In recent measurements we and Giffard 4 studied the thermometric properties of powdered CMN using a noise thermometer as a standard. In the present work we have measured the 16-Hz susceptibility (both x' and x") of CDP against the same noise thermometer in the same experimental arrangement using pure 3 He as the thermal contact agent over the noise temperature range from 2.2 to 20 mK. Hence we can measure the magnetic and thermometric properties of powdered CDP and then compare them with those obtained for CMN.The demagnetization cell used to obtain the measurements is shown in Fig. 1. It contained 10.7 g powdered CMN as cooling salt and 2.7 cm 3 liquid 3 He at low pressure. The noise thermometer and CDP sample are magnetically decoupled from the cooling salt by a niobium shield but thermally connected to it via a column of pure 3 He.Although noise thermometry based on a Josephson-effect device is a relatively new technique, it has been used by Kamper et al. 5 down to 23 mK and by Giffard and us 4 down to 2.2 mK. We use an improved version of a method described earlier. 6 The noise-temperature sensor was fabricated from copper foil 4 mm wide, 15 mm long, and 0.025 mm thick. Its resistance was observed in a separate experiment to change less than 0.4% from 3 mK to 4.2 K. The measured noise temperature is greater than the actual Johnson noise temperature Tj by an amount called the "device temperature" and determined in this experiment to be 0.05 mK. We believe that the noise temperature, corrected for device te...