Recent measurements of the specific heat 1 ' 2 and the thermal conductivity 3 of liquid He 3 at low pressures have demonstrated, for temperatures less than approximately 0.04°K, the temperature dependences predicted by the Fermi liquid theory of Landau 4 and Abrikosov and Khalatnikoy. 5 However, the magnetization self-diffusion coefficient measured using the spin-echo method by Anderson, Hart, and Wheatley 6 (AHW) obeyed a T-* 2 temperature dependence down to 0.03°K in contradiction with the T" 2 dependence predicted by Hone 7 on the basis of the Fermi liquid theory. As pointed out by AHW, their experiment was susceptible to error, particularly at the lowest temperatures, because of long equilibrium times and difficulties with thermometry. Consequently, the experiment was repeated under much improved conditions and extended to lower temperatures. We now find the diffusion coefficient to be proportional to T" 2 below 0.04°K -0.05°K. The nuclear susceptibility is constant below 0.1°K and equal to the value previously measured (AHW) within experimental accuracy. The lowest temperature of these measurements was 0.02°K. The measurements were carried out at a pressure of 12 cm Hg.The cell containing the He 3 was made from Epibond 100 A, 8 and the plug which closed the cell was sealed with Epibond 121. 8 Both resins are nonmagnetic. 9 The space containing He 3 consisted of a cylinder 1.2 cm high and 1.2 cm in diameter with a coaxial cylindrical tail 5.1 cm long and 0.25 cm in diameter. The inside surface of the cell was lined with 400, 0.004-inch diameter, 99.999% pure annealed copper wires about half of which did not extend into the tail. The larger cylinder contained 1.29 g of powdered cerium magnesium nitrate (CMN) which served as a thermometer. The CMN was prevented from entering the tail by a cotton plug. The He 3 was introduced into the cell by a 1/64-in. o.d. by 0.003-in. wall 70-30 cupro-nickel tube which had a negligible magnetic effect. 9 A pi-wound, 3000-turn pickup coil of 0.002-in. diameter manganin wire was placed over 2.0 cm of the tail at the greatest possible distance from the CMN. The above construction gave internal equilibrium times of 15 minutes or less at the lowest temperatures while providing sufficient separation between the He 3 being meas-ured and the CMN to give a negligible temperaturedependent magnetic gradient. The He 3 was cooled by 37.8 g of potassium chrome alum using techniques described elsewhere. 10 The thermometer was calibrated in the temperature range of 1°K to 0.3°K using a 470-ohm Speer resistor which previously had been calibrated in this temperature range with a CMN magnetic thermometer. The accuracy of the temperature scale is estimated to be 2 %.The He 3 used had a He 4 content of less than 3 parts in 10 5 and a tritium decay rate of less than 900 disintegrations/sec. The apparatus outside the cryostat and the method used to obtain the echoes were similar to those described elsewhere. 11 The self-diffusion coefficient, D, was measured by the method of spin echoes using an ...

No abstract

No abstract

No abstract

No abstract