This paper has prompted me to supply some unpublished information and to make several comments. The paper (Roman and Sermon, 1934) mentioned by these authors described a magnetic gradiometer designed and constructed at Michigan College of Mining and Technology (now Michigan Technological University) at Houghton, Michigan, in 1933. At that time, we were not familiar with Haalck's paper of 1925 nor with the previous work mentioned below. As stated, materials and design details were based on supplies locally available, and the development was to test the feasibility of using paired earth inductors to measure directly the earth's magnetic gradient. Field tests indicated that the instrument held promise and justified plans for improvement. Readings seemed reliable and repeatable, and the instrument proved to be more sensitive than had been expected.