Interested by William Marston’s studies in how a laboratory subject’s attempt at deception effects blood pressure changes using the discontinuous method, and encouraged by Berkeley Police Chief August Vollmer, John Larson attempted to use a deception test in actual criminal investigations. He examined a variety of instruments that were available at the time, as well as methodology. He selected an Erlanger Sphygmomanometer modified to produce permanent recordings of blood pressure using a smoked drum and kymograph. In his subsequent publications and reports on his experiments, John Larson was careful not to call this new "apparatus" a lie detector. He pointed out that his experiments were being conducted to develop a methodology of detecting deception rather than to "invent" a lie detector machine. But historians were to drown out this cautious approach to his work. History will record that John Larson developed the first polygraph instrument.

Contrary to popular opinion, John Larson utilized two separate instrument designs in his early experiments in the detection of deception

His first apparatus, he referred to as a "Cardio-Pneumo Psychogram," consisted of a modification of an Erlanger Sphygmomanometer. The modification was done by Earl Bryant for Dr. Robert Gesell of the Department of Physiology of the University of California. The first instrument Larson used in his experiments was borrowed from Dr. Gesell.

In 1921, Earl Bryant made another instrument for John Larson, which was used at the Berkeley Police Department. It is believed that this second instrument was a duplicate of the first. This is the instrument that brought fame to Larson’s experiments and that drew the young Leonarde Keeler into the field of detection of deception. He used a breadboard as a base for the instrument, and from that it became known in the industry as the "Breadboard Polygrah."

His instrument was loosely fashioned after Marston’s polygraph. Like Marston’s instrument, it recorded respiration and cardio activity. His instrument differed from Marston’s in that it provided continuous readings, rather than discontinuous readings. He also replaced the blood pressure technique used by Marston with an occlusion Sphygmomanometer Plethysmograph that measured relative blood pressure and blood volume. The instrument was therefore able to continually and simultaneously record respiration, and cardiovascular changes.

It had many drawbacks. It took a half hour to set up. The paper had to be smoked and was smudgy and messy. To preserve the graphs, they had to be shellacked and stored in cans and even then they often became brittle and broke. But it worked well and what he and Leonarde Keeler learned about instrumentation using this first "lie detector"to solve crimes in and around California between 1921 and 1935 formed the basis for future polygraph designs.

Described by Eloise Keeler as a "strange contraption of tubes, wires, a glass bulb, and a wide strip of black smoked paper which moved on two wooden cylinders. Two metal styluses or needles were suspended close to the paper." The pens scratched physiologic tracings into the smoked paper.

Over the next fifteen years, John Larson collected hundreds of files on successful criminal cases where his Cardio-Pneumo Psychograph solved murders, robberies, thefts and sex crimes. His instrument was nicknamed "Sphyggy" by the press who covered Larson’s crime solving escapades in the 1920’s and 30's. They couldn’t pronounce "Sphygmomanometer," hence "Sphyggy."

This "first" polygraph instrument is now at the Smithsonian Institute in Washington D.C.

Later, after moving to Chicago, Earl Bryant made the second modification to this first instrument by replacing the smoked drum as a recording device with a Jacquet Polygraph to record the tracings on paper and ink. Professor H.W. Edwards, a physicist in the University of Southern California described Dr. Larson’s second instrument as:

The apparatus consists of the rubber sleeve, a pressure reducing device, Marey’s tambour, a light glass lever and stylus, a roll of smooth pater and a kymograph or a motor driven drum for winding up the paper. The instrument is so designed that it is peculiarly sensitive to small changes in blood pressure.

The glass lever pen was drawn from 1/4 inch glass tubing so that the smaller end was about 0.25 mm in diameter. It is fastened to a steel pivot near the larger end. Motion was in a horizontal plane. To the smaller end was fastened a fine pointed piece fo bamboo which served as the inking point. Adjusting screws served to control the pressure of the pen point upon the paper. Any good fountain pen ink was found to be satisfactory. Very thin rubber was used on the tambour and this was allowed to be somewhat loose, not tightly stretched. The moment of inertia of the whole inking system was as small as could be easily made so that a more exact record of the pressure variation could be made than is possible with the heavier steel levers.

On very commendable feature of the apparatus was the fact that the time record was not limited to a few minutes. Records were made continuously for 20 minutes or more if desired. With the apparatus constructed for experimental work in this laboratory, it was found convenient to make simultaneous records of the blood pressure, respiration, time, and signal upon a four inch strip of paper.

In 1932, Chester W. Darrow, of the Institute for Juvenile Research, made a third modification to the Larson Cardio-Pneumo Psychograph, by adding a galvanometer. The new instrument included a psycho-galvanometric record, electrodes on the palm and back of the hand, as well as a continuous blood pressure record, and a pneumographic record.

Little is known of this instrument.