A study was made of the development and fate of individual plants growing from contrasting seed sizes in swards of subterranean clover (var. Bacchus Marsh). Three sets of swards were grown in boxes, from large, small, and mixed (large and small) seeds respectively, all sown at the same number of plants per unit area; the swards were harvested on five occasions. Leaf area in each 2 cm layer of the sward was determined separately and measurements of the light intensity reaching each layer were obtained at each harvest. Counts of the plants on each occasion revealed a steady drop of about 30 per cent. over the whole period; in the "mixed" swards, only plants from small seed sizes died, the number of large-seeded plants remaining unchanged. The data of leaf area and numbers in each layer showed that leaf area is concentrated in a relatively shallow band at the top of the sward, the area attained by individual leaves increasing with height in the sward and with time. Leaves are not retained in the canopy at heights below that at which light extinction occurs. An unexpected finding was that, as the sward developed, the leaf area index (the ratio of leaf area to ground covered) required for the absorption of all incident light energy increases, and it is suggested that this is due to changes in leaf morphology. The importance of the position at which leaves of individual plants are held in the canopy is stressed; although at the first sampling occasion, the small-seeded plants had 25 per cent. of the leaf area of the mixed sward, they were so located as to absorb only 10 per cent. of the incident light energy; at the last sampling occasion, the values were 10 and 2 per cent. respectively. It is concluded that the disappearance of the small-seeded plants from the mixed swards was due to shading from the large-seeded plants.

SHORTER CONTRIBUTIONS 231moved to the front of the apparatus and its chamber supplied with a cover (not shown on the photograph). T h e pen arm is fitted on to a gate which allows the pressure exerted on the chart by the pen to be finely adjusted. DrscussroNThis apparatus was constructed with the limited object of determining the length of time leaves were wet, but it could be used for other purposes such as measuring the amount of dew deposited, which would be valuable to agricultural meteorologists, particularly in arid zones. By measuring the surface area of the shoot and calibrating the deflection of the beam, the weight of water per unit area can be estimated. In 1953, one shoot was measured, and it; surface area found to be 1,480 cm2. Rain to ' run-off ' added 14.2 g, or 9-6 x l W 3 g cm-2. The maximum increase in weight during the three dew nights while this shoot was exposed were 4.7 g, 7.0 g and 10.2 g. The weights of dew per unit of total surface area were 3.2,4.7 and 6.9 x g cm-2, and the rate of deposition ranged from 6 to 8 x lo-' g cm-* hr-I in the period between the start of dew formation and the time when it reached its greatest weight. Estimates of the mean weight of water over the total surface area of the shoot obscure the differences of distribution and persistence of water deposits at different levels, on the stems and petioles compared to the leaf surfaces, and on different parts of the leaf. These differences are probably very important to pathology; for instance, the frequent potato blight lesions on the tip of the leaflets are often associated with large persistent water droplets. They are more common after rain than after dew, which rarely approaches run-off under English conditions. Interesting observations might be made on these differences by the use of several balances bearing variously dissected shoots. (11) on a monthly basis have been collected for 32 stations and regression equations relating the data calculated in the form Q = QA (a + bn/N) where QA is the mmimum possible ndiation in the absence of an atmosphere, and A! is the maximum possible duration of bright sunshine. For five of these stations records for long series of individual months have been examined and for all stations mean monthly values. In the overall regression equation a = 0.23 and b = 0.48. SUMMARY Records of solar radiation (Q) and of duration of sunshine

Changes in the pre-emergence distribution of dry matter in subterranean clover (Trifolium subterraneum L.) variety Bacchus Marsh were followed at 21°C, using three sizes of seed and three depths of sowing, ½, 1¼, and 2 in. Decreasing seed size and increasing depth of sowing both reduce the weight of the cotyledons a t emergence. Seed of the three sizes were sown a t three depths in pot culture a t staggered intervals so that emergence was simultaneous. Dry weight in the early vegetative stage was proportional to seed size, and total leaf area and leaf numbers showed similar trends. Plants of each seed size grew at the same relative rate. No effect of depth of sowing could be detected, and this was shown to be due to the cotyledon area a t emergence being constant for any given seed size, regardless of varying depth of sowing and hence of cotyledon weight. It was concluded that seed size in a plant having epigeal germination and without endosperm is of importance: firstly, in limiting the maximum hypocotyl elongation and hence depth of sowing, and secondly, in determining cotyledon area. Cotyledon area in turn influences seedling growth, which is not affected by cotyledon weight. Once emergence has taken place, cotyledonary reserves are of no further significance in the growth of the plants.

An experiment was carried out which enabled a separate assessment to be made of the influence of, firstly, pre-emergence growth rate, and secondly, the competitive environment at emergence on the subsequent growth of individual plants of subterranean clover. Swards of the Bacchus Marsh variety were grown in large seed boxes at the Waite Agricultural Research Institute, Adelaide. Four different sowing treatments were applied whereby the sowing of alternate plants in a checkerboard grid was delayed by 0, 2, 4, and 8 days. Emergence date and final dry weight were recorded individually for each plant in the experimental swards. There was considerable overlap in emergence times between the earlier and the later-sown plants. Consequently by a comparison of the separate regressions for earlier and later-sown plants of yield per plant on day of emergence for each treatment, it could be demonstrated that plants with very different pre-emergence growth rates but emerging on the same day in the same sward, and hence into the same competitive environment, gave essentially the same subsequent growth. In other words, in the absence of differential competitive effects, there was no detectable influence of pre-emergence growth rate on subsequent growth. On the other hand, the final dry weights of individual plants were very strongly related to day of emergence. A delay in emergence of 5 days brought about a reduction in final yield of about 50%, and a delay of 8 or 9 days of at least 75%. In view of the negligible influence of pre-emergence growth rate, per se, this reduction is attributable solely to the increased severity of competition from plants already emerged. In spite of the wide variation in emergence patterns, the total yields of all swards were very similar, indicating an almost complete compensation between plants in the utilization of available growth factors.

No abstract