When was the linearizing coil introduced?

Looking at the patents:

SU890 (Russia, filing date 06/23/1921) – oscillator without a linearizing coil,
US1661058A (USA, filing date 12/05/1925) – oscillator without a linearizing coil.

In RCA and custom theremins for Clara and Lucie, the linearizing coil is already present.

It would be interesting to know when and under what circumstances Lev Thermen came up with this idea. In principle, it seems to merit its own separate patent.

"US1661058A (USA, filing date 12/05/1925) – oscillator without a linearizing coil."  - ILYA

Oh, so interesting to read that, thanks ILYA!  In it he's aiming for 500kHz to keep it out of the AM band, I'm going for 450kHz for similar reasons.  He also refers to the field as "electrostatic" which I feel is a misnomer since the field is AC.  It also seems he's buffering the oscillators before mixing, shades of the ESPE01?  Also mentioned are multiple pitch antennas for a polyphonic instrument.

I wish I could shed some light on when the linearizing coil was introduced, but I got nothing.  I'm guessing he sensed the non-linearity of the near field and decided a double resonance might help ameliorate it to some degree, and went down that path.  It seems he's almost part of the way there with the transformer coupling to the antenna.  I agree that it probably deserves a patent of its own.

How does one view SU890?  I couldn't find it.

I've uploaded the Russian version here (SU890rus.pdf).

I no longer remember where I downloaded it from. There used to be an excellent website, trmvox.ru, hosting a vast collection of documents and articles from various magazines. I managed to download quite a few of them before it went down.

Please note that in many publications about Lev Termen (Theremin), this patent is incorrectly cited as №780.

Here (SU890eng.pdf) is the machine translation into English.

"I no longer remember"

It seems it was the official website of Rospatent (the Russian Federal Service for Intellectual Property).
However, it now only displays a reference to the patent, without its full text:
https://www1.fips.ru/iiss/document.xhtml?faces-redirect=true&id=f10c1586e56c930be35eaef170880d19#

This documents are indeed очень интересно, thank to ILYA. 

I suspect, because of the principle of an lengthen coil for oszillators and antennas was always known that time, Lev Termen used it from the beginning of his invention. Without this was a frequency change of some kHz by hand-capacitance with the early tube technique not possible. And without this possibility the instrument not of interest for anyone. 

"Without this was a frequency change of some kHz by hand-capacitance with the early tube technique not possible."- JPascal

This is a bit of a surprise to me, though I'm not a tube guy.  What is the limiting technology here?

"Without this was a frequency change of some kHz by hand-capacitance with the early tube technique not possible."- JPascal This is a bit of a surprise to me, though I'm not a tube guy.  What is the limiting technology here? - dewster

I’m not particularly familiar with the early triodes from the 1920th. However, there are some limitations when using Hartley, Coolpitch, or Meissner oscillators with the low capacitances required for a theremin. These are:
- a transconductance below 1, typically 0.5 mA/V;
- low gain and thus strong feedback to compensate for losses; and a grid leakage current.

Most important, however I meen, were the high capacitances between the grid and anode, as well as between the grid and cathode. The were further enhanced by the so called Miller effect. Thus, good stability was achieved for an oscillator in the 300-kHz range using capacitors of a few hundred pF. 

Because of Lev Termen found the for audio useful hand capacitance effect of the beat oszillations accidently while investigate the conditions for wide distance and approximation sensors it is near to assume, that there were high inductiv lengthen coils in use, which were coupled with a stable oscillator design.