104 The Essex Naturalist At Wake Arms tracks were abundant near the road edge. This is an area very little visited, as the forest is divided into small wedges by the roads radiating from the busy roundabout, and there are no rides or tracks down which it is convenient to walk. There is access to agricultutal land outside the forest at Stable Shaw, but this is 1000 m from the deer at Sunshine Plain, Deershelter Plain and Jack's Hill. Thus some deer signs are beyond our 750 m line and is the kind of area where deer should be found (according to the historical hypothesis). Then we come to the counts for the Lower Forest. If the historical hypothesis is correct, deer numbers should be lower than at Wake Arms, as it is more visited than the Wake Arms area being closer to habitation. The official count for the Lower Forest is 25 in 1996, and 11 for 1997. The keeper's estimate is larger, but unfortunately not proven. This seems quite a large number for a small area, larger perhaps than the historical hypothesis would predict. The deer are mainly located on the extreme forest edge, however, which is an appropriate place for them to be in terms of visitor disturbance as few people choose to cycle or walk their dogs there. The key factor here is where they are feeding. According to the historical hypothesis, they should be feeding solely in the forest, but the evidence of the slots discussed in the behavioural hypothesis fieldwork shows that this is not what is happening. They are crossing to the fields on the west to feed, and returning in the early hours of the morning. A point against the historical hypothesis. Visitor numbers are next and the number visiting the Epping Information Centre is pertinent to this enquiry. These are people who have come specifically to see the forest and find out more about it. Very many more visit the forest in a year including those who regularly use the forest for dogwalking, mountain biking and horse riding. The evidence is that numbers of visitors to the forest is increasing and the car park evidence supports the idea of greater visitor impact to the south of the main forest. These facts which correlate with lower deer numbers in the south are a point in favour of the historical hypothesis. The forage nitrogen analysis is preliminary but suggests that nitrogen content for the grass within the forest is well below that of the crops and grass in the agricultutal fields which receive fertiliser. It is known that there is a real deposition of nitrates on to the forest, for example, from M.25 traffic emissions. However, it appears from initial measurements here that better grazing is available outside the forest. "Deerlawns" within the forest are unlikely to be sufficient in area and quality to support the deer population. A point against the historical hypothesis. The final piece of fieldwork for this hypothesis concerned coat colour. Common fallow numbers far exceeded black fallow numbers in the bufferlands. This does not agree with the historical idea that the large numbers of fallow on the bufferlands are due to black coated 'forest' deer being pushed north. Two black deer cannot produce a common fawn (Bignell, 1993). Two common deer can produce a fawn of any colour. Thus how many of the black deer on the