On an acid, dry infertile soil in a frost pocket, there are more short shoots (potential fronds) in bracken than differentiated fronds, and more differentiated fronds than emerge from the soil, and more emergent fronds than live in August.Factors affecting the inception, differentiation and fate of fronds are examined. The inherent tendency of fronds from young shoots to develop and to emerge before those from old renders them more liable to lethal winter and spring frosts and to crippling spring frosts. Drought acts mainly on the late emerging fronds. The normal sequence of emergence may be upset.Through their destruction of reserves and their effect on the means of replenishing them, annually fluctuating frost and drought influence the number of shoots initiated yearly, and hence the age-structure of the shoot population. A senescent population of bracken is deflned as one in which the proportion of young to old shoots is lovi-, and fronds from young shoots to fronds from old is also low. Frost and drought are the causes (not necessarily exclusive) of senescence in this area.The validity of the generalizations by Braun (1851) and by Hofmeister (1862) on the number of fronds produced per annum on the shoot of the mature plant, and the duration of the period from the first visible appearance of the frond to its emergence from the soil, have already been called in question (Watt, 1940). Further evidence on these matters and on departures from the 'normal' behaviour of bracken is provided by data from a plot on Lakenheath Warren, on a very acid podsolized sand (pH 3.6-4.0) (area E in Watt, 1940).
METHODIn a marked plot 2 x 2 ft (0.37 m^) the fronds, on emergence or at convenient times thereafter, were charted (Fig. i) and identified by numbered rings in three successive years, 1956, 1957 and 1958. The plot was then dug up on 7 August 1958 and the shoots and their number of nodes counted. In this technique, despite the care exercised, there is the possibility of mistakes in identification, where marked fronds, killed by frost whilst still young, then decayed and left the ring above the remains but separated from them. There is also the purely physical difficulty of disentangling from the grass-rootpermeated soil the ultimate shoots without detaching them from their parent short shoots.
RESULTSThe data in Table i substantiate earlier findings (Watt, 1950): the small number of live fronds in August (16.7 per 10 ft^: 18.0 per m^-numbers which are close to the mean 75