Tn an agronomic experiment, perennial ryegrass was grown with four rates of fertiliser N + K at 21 sites in England and Wales. Herbage from two successive harvests at 14 of the sites has been examined for contents of N and S and of other mineral elements in relation to the nutrition of both the crop and ruminant animals. The highest rate of fertiliser application (125 kg N + 77.5 kg K/ha harvest) resulted in contents of total N greater than 4 % at five sites at the second harvest, but nitrate-N reached levels that might be toxic to livestock ( > 0.35 %) at only two sites. There was no evidence that supplies of S from soil and the atmosphere were inadequate for crop growth, or that the contents in the herbage were insufficient for ruminant animals, even though no S was supplied as feitiliser. On the contrary, in many samples, the content of S was so high as to constitute a potential hazard for ruminant animals through impairment of Cu availability. Values for the ratio total N : total S, which ranged from 3.2 to 15.6 : 1, generally increased with increasing rates of fertiliser. The concentrations of sulphate-S in the herbage were generally high and represented between 19 and 78 % of the total S. The content of Mg was only slightly affected by fertiliser N + K, but the ratio K : (Ca + Mg) increased markedly with increasing rates of fertiliser. At several sites the ratio exceeded 2.2 : 1, the value above which there is an increasing risk of hypomagnesaemia in ruminant animals. The sum of the concentrations of the cations C a + M g + K + N a in the herbage increased as the concentration of N increased, and there was a similar though less marked relationship between organic anions and total N in the herbage.

Non-volatile organic acids have been determined in S24 ryegrass and S37 cocksfoot harvested on three cutting dates over the period of ear emergence and in eight grasses harvested on two cutting dates. In both S24 ryegrass and S37 cocksfoot the concentrations of shikimic, quinic and total acids decreased with increased age of the plant. Correlation coefficients are given between the concentrations of acids in eight grasses and the 'preference ranking' of the grasses estimated in an investigation into the grazing preferences of sheep and cattle, and a regression is given for relating 'preference ranking' to the contents of citric and quinic acids. IntroductionMeasurements of the organic acids in herbage indicate that they form a n appreciable part of the dry matter; Fergusonl found that the total organic acids, calculated as malic acid, ranged from 4% to 7% on a dry matter basis; the method he used was t o determine the total acids extracted by ether. Organic acid contents of 5% to 6.3% were found in hays when an aqueous extract was used.2 Again with an aqueous extraction, Fauconneau & Jarrige3 obtained values of 3 . 5 O / , to 5% in herbage. More recently, methods that have been developed for separating the individual acids and measuring them use ion-exchange chromatography4, 5 or partition chromatography on silica ge1.6, Information about the changes in the non-volatile organic acid contents of herbage was required in connexion with work on herbage digestibility.8 Three cuts each of S24 ryegrass and S37 cocksfoot, harvested over the period of ear emergence, were examined. The concentrations of the non-volatile organic acids were also determined in eight different grasses on two occasions. These latter samples were taken from an 322Jones & Barnes: Non-Volatile Organic Acids of Grasses investigation into the grazing preferences of sheep and ~a t t l e ,~ the results from which had suggested that some constituent of the water-soluble fraction of herbage might have a better correlation with preference ranking than any of the other chemical components which had been measured.Ion-exchange chromatography was used for the measurement of the non-volatile organic acids reported in this paper. Because this technique for the determination of these acids was lengthy, it was decided that this approach would not be continued in further investigations on the digestibility and 'palatability' of forage grasses. However it was thought desirable to place on record the results obtained in this preliminary study. Experimental SamplingHerbage samples were harvested, dried rapidly at 100" in a forced-air oven, ground to pass a 0.8 mm sieve and stored in wax cartons. ExtractionFour methods of extracting the acids were investigated: (a) Ether e~t r a c t i o n ;~ (b) Hot-water e~t r a c t i o n ;~ (c) Acetone extraction followed by extraction with 0.01 N-HC~;~O (d) extraction with 80-85% ethanol under reduced pre~sure.~ Extraction from herbage with 80-85% ethanol gave lower values of citric and malic acids compared with the other three me...

Perennial ryegrass (Lolium perenne L. cv. S23) and white clover (Trifolium repens L. cv. Blanca) were grown in mixed swards, with and without fertiliser N as ammonium nitrate, at 12 sites in the UK. Grass and clover separated from herbage harvested in May, July and September, have been examined for N, S and other major mineral elements, in reIation to the nutrition of both the crop and ruminant animals. The total N content was greater in the clover than in ryegrass at each site, both with and without fertiliser N. The S content of the clover was similar to that of ryegrass at the first harvest in May, but was lower in July and September. There were large differences between the species in the ratio N:S, the average being 17.1 :1 in white clover and 8.6:l in ryegrass. The proportions of the total S present as sulphate were higher in ryegrass, with a mean of 48 %, than in the clover, with a mean of 16%. The contents of Mg, and particularly of Ca, were higher in clover, while K was generally higher in ryegrass. Consequently, the ratio K: (Ca+Mg) was always at least twice as high in the ryegrass as in the clover. In some of the ryegrass, but none of the clover samples, it exceeded 2.2:l. the value above which there is an increasing risk of hypomagnesaemia in ruminant animals. The content of total cations (Ca+ Mg+ K + Na) was on average 37% greater in the clover than in the ryegrass, while the content of total inorganic anions (but including organic P) was much less in the clover. The content of organic anions, as calculated from the difference between the contents of total cations and inorganic anions, was at least twice as great in white clover as in ryegrass.

Procedures based on five methods of soil analysis for predicting potentially mineralisable N, and hence soil N supply, have been evaluated and compared with prediction based on total soil N. The effect of adjusting the analytical results for differences in soil temperature and water status between sites has also been examined. The evaluation was performed by calculating the extent to which the results accounted for variation in the amounts of N in the herbage of perennial ryegrass plots receiving no fertiliser N. The plots were situated at 18 sites throughout the UK: at some sites results were available for more than 1 year, giving a total of 29 site/years. The best prediction, accounting for 65 % of the variation in herbage N yield, was obtained when the amounts of 'glucose' extracted from the soils by 0 . 0 5~ Ba(OH)2 were combined with the amounts of Nos-N extracted by 2 M KCI, and the combined values adjusted for soil temperature and water status. A procedure based on the measurerrent of the NH4-N released by autoclaving the soils with 0 . 0 1~ CaClz accounted for 51 % of the variation However, procedures based on (a) NH4-N extracted by acid KMn04, (b) non-NO3-N extracted by 0 . 0 1~ NaHC03, and on (c) measurement of the absorbance at 260 nm of NaHC03 extracts, were no more effective than was analysis for total soil N (less than 36% of the variation accounted for when values were adjusted for soil temperature and water status).

Scatter corrections are commonly applied to refine near‐infrared (NIR) spectra. The aim of this study is to assess the impact of measurement errors when using ordinary least squares (OLS) for multiplicative scatter correction (MSC). Any measurement errors attached to the set‐mean spectrum may attenuate the OLS slope and that in turn will affect the estimate of the intercept and the adjustment of the spectra when using MSC methods to mitigate scattering. A corrected least squares slope may be used instead to prevent this problem, although the impact of this approach on the final outcome will depend on the relative size of the measurement errors in the individual spectra and the set‐mean spectrum. The errors‐in‐variables or type II regression model (also known as Deming regression) and its special cases, major axis (MA) and reduced major axis (RMA), are discussed and illustrated. The extent of OLS slope bias or attenuation is demonstrated as is the resulting MSC spectral distortion. Further modification to the MSC transformation method is also suggested. The influence of scattering correction (by MSC, standard normal variate (SNV) and detrending) and of using the maximum likelihood estimate of the slope for MSC on the prediction of chemical composition of Lucerne herbage from NIR spectra was assessed. The predictive performance was slightly improved by the use of scattering corrections with fairly minor differences among methods. Nonetheless, it seems well worth considering the use of type II regression models for assessing MSC application aiming at improving the goodness of prediction from NIR spectra.