A curving stretch of dyke comprising ditch and rampart (on the estern or inner side) with western ditch built, c.2.4km long. Despite being one of the most excavated and observed stretches there is still no firm dating evidence. It is thought to be the earliest dyke if the whole dyke system on the basis of its topography and the fact that it is cut by Bluebottle Grove / Lexden Dyke. Lexden Dyke having been tentatively dated to Sheepen 1 period (AD 5-43). The relationship between Heath Farm Dyke and Shrub End Dyke is not clear, recent work suggests that Shrub End Dyke did not continue beyond Heath Farm Dyke and may have stopped at Kidman's Dyke North.

Henry Laver published a section drawing showing the profiles of Heath Farm Dyke Middle and adjacent Kidman's Dyke Middle c.1900.<1>

During the construction of a tank trap during WWII (in 1940 & 1943) A.F. Hall recorded that the southern Bluebottle Grove's ditch was seen to overly the ditch of Heath Farm Dyke.<2>

A section was cut across Kidman's Dyke North and Heath Farm Dyke Middle by A.F. Hall in the 1950's. The exact location of the section is not known, no further information.<3>

In 1957 C.F.C Hawkes sectioned Kidman's Dyke / Heath Farm Dyke. One of these is recorded as upstanding, presumably Kidman's Dyke. No further details.<4>

In 1956/7 C. F. C. Hawkes and A.F. Hall excavated sections through Heath Farm Dyke Middle just south of the junction with Bluebottle Grove Dyke and Prettygate Dyke. Here a gap in the dyke has been interpreted as a likely entrance. The section revealed that the ditch had a shallow rounded profile 1.5m in depth. The end of Heath Farm Dyke Middle ditch was surmised to be overlain by Prettygate Dyke which itself had been slighted for possible use as a Roman Road. This could not be tested because the entrance was under a public footpath.<5>

In 1957 three trenches were excavated by C.F.C. Hawkes across the dyke (including one across an entrance/gap) on the site of the then Co-operative Societies sports field TL 97022376 (now the area of Regency Green, off Kind Harold Road). Trench 1 revealed the (loamy) base of the rampart, c.150mm thick. Trench 2 recorded the rampart width as 7.3m. The profile and fill of both trenches were similar to the later 1961 excavation, although the 1957 section looked shallower and more rounded. The third trench revealed an entrance causeway, c.3m wide.<6>

In 1959 A.F. Hall recorded the ditch of the dyke at four points to the north-east of the Co-op Sports field entrance during building work for a new housing estate, confirming its line up until 'Prettygate Junction'. The profiles were similar to previous sections.<7> Unlocated

In 1961 C.F.C. Hawkes cut through the ditch 116m south-west of Lexden Straight Road at Heath Farm. The base of the ditch was 1.8m from the farmyard surface, the lowest 180mm were very sandy gravel silt, beneath 80mm of darker sandy gravel silt, beneath 80mm of darker sandy old turf. On this lay the modern filling of stony red gravel thrown down from the rampart and covered by surface loam.<8>

The dyke ditch was observed in a sewer trench by J. Sewter in 1974. The oblique angle of the trench made interpretation difficult however the possible base of the bank appeared to survive to a height of 0.5m and was about 7m wide.<9>

Aerial photograph's taken by Ida McMaster during the 1970's show the dyke abutting the Roman auxiliary fort at TL963227 with uses the dyke as its western boundary.

The ditch of HFDM was visible in the western side of Shrub End gravel pit was cleaned and photographed. It was distinctly V-shaped and was 2.4m deep and 7.8m wide. The ditch had been backfilled with reddish brown sand and gravel, and to a depth of 0.4m with brown sandy loam. Recorded by J. Sewter in 1974.<10>

In 1984, Carl Crossan recorded an evaluation trench at 274-298 Straight Road, which revealed a shallow ditch 1.5m from the surface of the field. The rampart was not well defined.<11>

A watching brief at Shrub End Landfill site noted that a bank, previously interpreted as part of the dyke, appeared to be a bund of soil created next to the property boundaries of the adjoining residential estate. The soil was a light brown-yellow brick earth material displaying much root disturbance, and otherwise sterile. It is suggested that the dyke had been quarried away at this point. Recorded by S. Gibson in 1995.<12>