Waveguide, horn antenna and semiconductor detector

Plumbing for invisible light

Bose wanted to show that radio waves behave like light, so he needed waves short enough to bend, focus and polarise on a tabletop. His transmitter was a spark gap mounted inside a metal tube, which in modern terms is an overmoded circular waveguide: the tube confined the radiation and sent it out of one end ²⁴. In front of it he put lenses of glass or sulphur to collimate the beam, the first waveguide-lens antenna ⁴.

On the receiving side he used what he called a "collecting funnel", a pyramidal metal horn that gathered the waves into the detector. Emerson credits him as the first to use an electromagnetic horn antenna ⁴. He also built prisms, a cylindrical diffraction grating and polarisers, one made from a railway timetable interleaved with tinfoil ⁴⁶. He presented this apparatus at the Royal Institution in London in January 1897, working at wavelengths from 2.5 cm down to 5 mm ⁴.

A crystal and a micrometer screw

The hardest part was the detector. Coherers of the time, tubes of metal filings that clump when a wave arrives, had to be tapped to reset. Bose experimented with many materials and found that a fine metal point pressed against a crystal of galena (lead sulphide) responded to radio waves and recovered by itself ⁴⁵. His US patent, filed on 30 September 1901 and granted on 29 March 1904, describes two contacts held together by a spring, with a micrometer screw to set the pressure "with great delicacy"; it names galena as a suitable material, and notes that fitted with a lens the device also detects light ⁵.

This was a point-contact semiconductor detector, the principle behind the cat's-whisker crystal radio sets that followed. When Pearson and Brattain of Bell Labs wrote a history of semiconductor research in 1955, they credited Bose with the first use of a semiconductor crystal as a detector of radio waves ⁴.

Sixty years ahead, or a precursor?

Nevill Mott, who shared the 1977 Nobel Prize in Physics for work on solid-state electronics, said that "J.C. Bose was at least 60 years ahead of his time", and that he had anticipated the existence of P-type and N-type semiconductors ⁴. In 1998 the Proceedings of the IEEE marked the centenary of the semiconductor diode detector and reprinted Bose's patent and papers ⁶.

Historians add a caution. Kapil Subramanian points out that Bose's materials work is cited as a precursor to solid-state physics, but that he did not discover rectification, and that a radio-telegraphy company once asked him to hold back publication until a patent deal could be negotiated ⁷. Bose patented only this one device and did not develop it commercially, so the crystal detector and microwave horn were developed again by others in the West, and his priority was recognised only decades later ⁴⁷.

Still in every microwave lab

The IEEE Milestone dedicated at Presidency College in 2012 lists Bose's spark transmitter, dielectric lens, polariser, horn antenna and diffraction grating, and notes that modernised versions of these are used today in Earth-to-space links and remote sensing ⁶. Horn antennas feed satellite dishes and radio telescopes; waveguides carry power in radar; and the semiconductor diode, born as a crystal and a spring, is in every electronic device ⁴⁶.