The method suggested by Papadakis of using covariance with the yields of neighbouring plots to reduce the error of replicated field experiments is illustrated on two large-scale cotton experiments. In a discussion on the validity and value of the method, it is concluded that for such experiments, where the number of plots per block is large, the method should be approximately valid and sometimes useful.

A method of obtaining samples of semen from rabbit males by means of a dummy and artificial vagina is described. The volume of the semen, and the number of spermatozoa per cm. (density) and the total number of spermatozoa in the sample have been measured. The data obtained from these measurements have been analysed under the following headings:

1. Relation between total volume and total number of spermatozoa. With an increase in the total volume of the ejaculate there is an increase in the total number of spermatozoa, but the relationship is not linear. The density of spermatozoa varies considerably.

2. Seasonal variation. Seasonal variation in sperm production is not marked in these data. There is slight tendency for the numbers to be highest in the early part of the year and lowest in August and September when the animal is moulting.

3. Individual variation. In each individual, variation is considerable. Differences between individuals are not marked, but as most of the animals were of low average fertility, it is probable that this result is not typical of male rabbits in general.

4. Effect of successive matings with in a short period. In successive matings the number of spermatozoa diminishes rapidly. As the density is not much affected, the diminution is due to reduction in the volume of the ejaculate.

5. Relation between the number of spermatozoa in the ejaculate and the number in the various sections of the spermatic tract. After a series of samples had been obtained, some males were killed and an estimate made of the number of spermatozoa present in various parts of the tract. From the results it is concluded that on ejaculation the vas deferens is practically emptied and that some sperm in the ejaculate come from the cauda epididymis.

6. Observations on sexual behaviour. The sexual pattern of the male rabbit is described and the relation between sexual stimulation and sex drive is discussed.

1. Extractions of soils with sodium hydroxide, followed by an acid, have been used in an attempt to fractionate the soil phosphorus.

2. Colorimetric methods for the estimation of the organic and inorganic phosphorus in alkali soil extracts have been suggested.

3. The amount of soil phosphorus soluble in sodium hydroxide is affected by the active soil calcium. It is suggested that sodium-saturated soils be used when studying the alkali-soluble phosphorus.

4. The acid-soluble phosphorus remaining in soil after extraction with sodium hydroxide was determined. This fraction appears by analogy to be similar to the apatites.

5. The largest fraction of the total soil phosphorus was not dissolved by the sodium hydroxide and acid extractions. This fraction was not increased by the long-continued use of phosphatic fertilizers at Rothamsted and Woburn.

6. Relatively large amounts of organic phosphorus were found in soils and the amounts were closely related to the carbon contents.

This publication should be read in conjunction with the paper immediately preceding it in the present issue of this Journal (Woodman & Evans, 1938). It deals with a particular phase of the general problem described in the first paper, to which reference should be made for details about the wider objects of the trials, the design of the experiments and the different manurial treatments to which the subplots were subjected.

1. Studies lead to the conclusion that Chibnall-preparations of pasturage proteins, in spite of their small amounts of carbohydrate impurity, may yet yield, on hydrolysis by mineral acids, practically the right amounts even of most of those amino-acids which are adversely affected by carbohydrates to some extent when tested singly under the conditions.

2. Preparations made in another way after treating the pasturage with ether as in the original Chibnall-method showed features of interest in the problem of attaining very pure proteins now much needed for purposes of comparison.

3. A new method is introduced by which it is possible to compare what may be extracted from the protoplasm of fresh unpulverized pasturages with the composition of the natural saps on a unit weight basis.

4. Production and breakdown of protoplasmic protein in perennial rye-grass during growth consistently appeared related to variations in the content of “nitrogen” and phosphates in the saps and the supplies of “nitrogen” and phosphates from the soil.

In continuation of earlier work, a study has been made of the factors controlling the competition between spring-sown cereals and annual weeds. During the years 1934 and 1935 some eighteen replicated trials were carried out in widely different localities. In six experiments, observations involving some 20,000 counts were made on the cereal development. Competition between Brassica arvensis (yellow charlock) and spring barley primarily reduced the number of tillers and fertile shoots; competition on the other hand with Raphanus raphanistrum (white charlock) diminished in addition ear size. The presence of R. raphanistrum in spring oats may decrease panicle size to the same extent as shoot number. In four out of six experiments the addition of nitrogen to both the weedy and clean crop had similar effects, greatly increasing tiller production in barley and both panicle size and shoot number in oats.

In six experiments nitrogenous manuring raised the yield of the weedy crop (containing either B. arvensis, R. raphanistrum or Chrysanthemum segetum (corn marigold)) to or above the level of the clean crop; n i four the yield increases were greater in the weedy than in the weed-free crop. In one experiment doubling the amount of nitrogen applied to the weedy crop did not increase the yield of oats more than the single rate, although the final yield level was considerably below that of the weed-free oats. In four other experiments additional nitrogen did not raise the yield of the weedy crop; but in only one was the yield of the clean crop increased by nitrogenous manuring.

This paper deals with the final phase of the investigations described in the two papers immediately preceding it in the present issue of this Journal. These earlier papers may therefore be consulted for complete details concerning the general design and wider issues of the work as a whole (Woodman & Evans, 1938a, b).

1. The vitamin A content of the colostrum of one hundred cows, kept under the same conditions of dietary and management, was found to range from 1181 to 35 I.U. per 100 ml.

2. It has been established that this wide range was not due to any differences in dietary, breed or date of calving in animals kept under the same conditions of management.

3. It is suggested that the vitamin A content of the colostrum may be materially affected by the length of the non-lactating period between successive calvings.

The main conclusions that can be drawn from these field experiments to compare the effect of different methods of preparing the seed bed on the crop's growth are:

(1) Crops germinate faster on the looser seed bed prepared by a Rototiller than on the more compact ones prepared by a plough or a grubber. The total number of plants that germinate is, however, the same for all treatments unless the land is too foul with weeds, when higher germination is obtained on the cleaner plots.

(2) Cereals tend to ripen a little sooner on land that has been ploughed than on land that has been either rototilled or grubbed, but in most years this effect if present is very small.

(3) The shape and the weight of the mangold root seems to depend on the seed bed. The roots were longest and thinnest on the deep-ploughed plots and were always squatter on the shallow-tilled than on the deep-tilled plots. The roots were heaviest on the deep-ploughed plots and lightest on the rototilled plots. On the rototilled and the grubbed plots the depth of tillage had no effect. The plants on the shallow-grubbed plots seemed, however, to have no reserve of strength, for they could not make better growth if given more room, while those on the deep-grubbed plots could make some use and those on the ploughed or rototilled plots appreciable use of extra space.

(4) Weeds tend to accumulate on the rototilled and the grubbed plots since neither grubbers nor rotary cultivators carrying tines mounted on a horizontal shaft can bury weeds and weed seeds in the way that the plough can. If the land is fairly clean and in good heart this probably does not matter for several years, but it prevents either of these from completely displacing the plough.

(5) A subsidiary result that emerged from these experiments is that if a thin crop is given a nitrogenous top dressing, the fertilizer may benefit the weeds more than the crop.

In a previous paper (Pizer, 1937) it has been shown that the physical state of composted horse manure is of great importance as regards the growth of the mushroom mycelium. A granular condition obtained by flocculating the manure with small amounts of calcium—0·5 part of calcium per 100 parts of dry compost, wt./wt. —is most suitable for rapid, vigorous growth, while a dispersed greasy condition arising through a shortage of calcium is very unfavourable and may prevent growth entirely. In both flocculated and dispersed composts, mycelial growth is made more vigorous by small additions of a soluble phosphate, equivalent to 0·031 g. phosphorus per 100 g. of dry compost. Further experiments dealing with the practical application of these results are described in the present paper. The calcium compounds used were commercial flake calcium chloride (26),1 ground gypsum (24), superphosphate of lime (21) and hydrated lime (53), the latter being used in error for ground carbonate of lime (38). All these materials, except hydrated lime, were found to be satisfactory in laboratory trials for promoting mycelial growth, but the best results were given by the first three. The materials differ greatly in composition and properties, in ease of application and reactivity, and probably in their influence on the metabolism and productivity of the mushroom, and it was the main object of the experiments to discover the most suitable one for the grower to use.

The fact that land which frequently bears a crop of clover becomes rapidly unfit to bear further crops of this plant is a very old observation in all countries where clover is grown. In England, Arthur Young (1804) called attention to it and to the fact that farmers were accustomed to call this unfitness “clover sickness”. The cause of this soil condition has been a matter of careful study since early in the nineteenth century, and some of the most interesting observations on the subject were made by Lawes & Gilbert (1860) in the years preceding 1860. They showed that the lack of capacity to grow clover did not seem to be connected with any deficiency in the principal specific plant foods in the soil, and the only contrary evidence was the fact that when a rich garden soil was substituted for the ordinary field soil, it was possible to grow clover for a much longer period without failure (Gilbert, 1871). The final conclusion of Lawes & Gilbert as to the so-called clover sickness was that the only means of ensuring a good crop of red clover is to allow some years to elapse before repeating the crop on the same land.

A recent paper by A. E. Brandt (1937) goes into the details of a type of design in field experimental procedure, where two or more factors are under examination, which has been much elaborated since 1928, when the first 2 × 2 × 2 experiment, involving two levels of each of three factors, nitrogen, potash and phosphate, was carried out at Woburn (Rothamst. Rep. 1927–8) with four-fold replication. Similar experiments, of the 3 × 2 × 2 type, had in fact been conducted at Rothamsted (Rep. 1925–6) two years earlier, but there was here the further complication that no differentiation was possible for two of the factors at one level (no manure) of the third. Details have been given by Fisher (1937) and Yates (1937) of, among others, experiments of the 2n and 3 × 2n types, and it may be said that the recent work has been in the direction of systematizing the lay-out and analysis of such experiments. Further features have been the device of confounding, which dates back to 1927 (Rothamst. Rep. 1927–8), i.e. it is almost contemporaneous with the first introduction by Fisher of randomized blocks and Latin square experiments, and the suggestion that replication may even be dispensed with entirely, a much more recent innovation. Confounding is a method of enlarging the number of blocks between which elimination of soil heterogeneity is possible by sacrificing information on certain of the higher-order interactions, which are considered unlikely to be real effects; with absence of replication an estimate of the experimental error is found by grouping together a number of these higher-order interactions.

Although it is well known that application of nitrogenous fertilizers to wheat often results in considerable increases in yields of grain and straw, and while experiments at Rothamsted and Cambridge have been very valuable not only in demonstrating the magnitude of the yield increases but also in showing how nitrogenous fertilizers bring about these increases, there is singularly little experimental evidence available regarding the increases in yield which may be expected in other parts of the country. Further, there is very little experimental evidence as to the optimum time of application of nitrogen, except at Rothamsted and Cambridge, both of which are in relatively dry districts.

1. Comparative analyses of pyrethrum flowers have been carried out by the methods of Tattersfield, Seil, Ripert, Haller & Acree and Wilcoxon.

2. The methods were of value in indicating the relative richness in pyrethrins of the samples tested, but discrepancies were seen in the absolute values of the pyrethrins I and II recorded. Under present conditions and until a standard method of analysis is agreed upon, it would appear requisite to state the method employed in the evaluation of the flowers.

3. The Wilcoxon method has given higher figures for the pyrethrin I content than the Seil method. The degree of divergence between the results depended upon the richness of the flowers, and upon the excess of acid used in distilling the volatile acid in the Seil method. The relationship between the amount of the pyrethrin I acid present and the titration recorded in the Wilcoxon method was not a linear one.

4. The question of the solvent to be used for the initial extraction of the flowers has been briefly discussed.

The literature which deals with the question of excretion of nitrogen by leguminous plants is reviewed especially with reference to accounts of experiments in which no excretion was obtained.

Experiments made at the Wisconsin experiment station prior to 1936, primarily concerned with the associated growth of pea and oat mixtures, were uniformly negative with respect to benefit to the oat plants through the association.

Experiments carried out at the Biochemical Institute in Helsinki, Finland, with a pea and barley mixture were in general positive, although poor development of the plants because of unfavourable weather were accompanied by little or no excretion.

On duplicating the Helsinki experiments at Wisconsin, once more both positive and negative results were obtained, although the development of the plants was excellent.

The occurrence of excretion appears to depend on many complicated circumstances some of which have been defined, while others are at present unknown. Factors which insure a uniform development of the plant cultures and especially a not too rapid growth of the leguminous plant appear to be important; as yet no universally successful method of achieving this type of development has been determined. It is concluded that although there can be no doubt with, regard to the existence of excretion, knowledge of the factors concerned is still insufficient to allow duplication of the results consistently at all experiment stations. Until such knowledge becomes available, application to practical agriculture can remain only an attractive possibility and not an established fact.

The experiments of Briggs & Shantz led them to conclude that the “wilting coefficient” is a soil “constant” which is (a) independent of the kind of plant used as indicator, (b) independent of the conditions under which the plant was grown, and (c) directly related to several other soil constants.

Subsequent research as well as an examination of their own results has shown that (c) is untrue, while (a) and (b) are substantially correct for hygrophytes and mesophytes. Earlier writers have been led to wrong conclusions regarding (a) and (b) through assuming (c) to be correct and through disregarding the particular nature of “permanent wilting” as defined by Briggs & Shantz.

The fact that considerable variation is to be found between the osmotic pressure found in different plants, in different parts of the same plant and in the same part under different conditions is not at variance with conclusions (a) and (b) when properly understood.

An important factor making for the substantial constancy of the “wilting coefficient” for a given soil is the extreme steepness of the curve connecting suction pressure and soil moisture content, in consequence of which differences of suction pressure of unquestionable significance from the standpoint of plant physiology give rise to differences in soil moisture content that are too small to be detected.

1. The high grain-yielding capacity and low total nitrogen content of predominant British varieties of barley, in conjunction with high ear survival at harvest, are discussed.

2. The results of a series of quantitative comparisons in which increased tillering was promoted by the application of nitrate of soda at various phases of plant development, and similarly when tillers were removed at various stages, were examined.

3. It was shown that the effect of nitrogen on the yield and quality of the grain depends upon the time in relation to the stage of plant development at which it was applied; the early applications enhanced the yield without detriment to the quality of the grain, whilst the later applications did not increase the yield but increased the total nitrogen content, and consequently reduced the malting quality of the grain.

4. The increase in yield was obtained primarily by an increase in the number of surviving tillers and to a less extent by an increase in yield of grain per ear; the latter was in some cases accompanied by an increase in 1000-grain weight. High ear survival in varieties such as those under examination may consequently be regarded as an index of high yield of grain and of low total nitrogen content.

5. The removal of tillers resulted in increased total nitrogen content and increased 1000-grain weight in the grain of the main stem; the former was not affected by the stage of development of the plant at which abscission was made, but the highest value of the latter was obtained when the abscission was made before flowering.

6. The relation of 1000-grain weight and nitrogen content was discussed, and it was shown that high 1000-grain weight may exist with both high and low total nitrogen. Although the 1000-grain weight is an hereditary attribute it is subject to considerable fluctuation and cannot by itself be usefully employed as an index of nitrogen content, and consequently of malting quality, in either an inter- or intra-varietal sense.

1. The proximate analysis has been made of several profiles from the north-east of Scotland, these including Scots pine, beech, birch and Calluna heath.

2. The system of proximate analysis proposed by Waksman has been used as a basis for the investigation, but various modifications have been introduced, e.g. sugar and cellulose determinations, with a view to obtaining greater accuracy.

3. Although a restricted number of profiles has been investigated, the results indicate that these may be divided into two groups, viz. raw humus and mull types, according to the manner in which the various fractions have been decomposed.

In the spring of 1937 an experiment was conducted to test existing commercial facilities for the transport of bull semen between England and Holland. Samples were sent to and received from Holland via Royal Dutch Air Lines, to whom the authors wish to express their sincere thanks for most useful co-operation. The minimum interval between collection in one country and insemination in the other was 28 hr. and the maximum, 57 hr.