response. The 2• rule and the 2:3:5 proportions employed in practice satisfy this criterion, but the present result suggests that a relatively broad range of proportions may be equally satisfactory. A quantitative evaluation of response irregu• larity, involving damping and other factors neglected here, remains for further study.The author is grateful to P.M. Morse for valuable advice and guidance, to H. Feshbach for helpful discussions and checking of the manuscript and to several colleagues for assistance as noted in the paper. APPENDIX I Probability Function for Random (Poisson) Distribution Consider a random distribution in which every normal frequency u• in an interval /x has an equal probability of occurring at any frequency in that interval. This is an example of Potsson distribution, for which it can be shown that is the probability that there are m normal frequencies in an interval in which N=zI/• is the expected (average) number. The probability of finding no normal frequencies (m=0) in the interval/x is Po=e-•r=exp(-/x/•). But if there are no normal frequencies in/x then there is a space of width /x. Therefore, exp(-/x/õ) is the probability orrincling a space of width/x where the average expected space width is •. The probability of occurrence of an actual space /5 is then exp(-/5/•)=e-ø; the probability per unit average space is (1/•)e-0; and the number per unit • in the range between/sand/5-1-d/5 is (1/•)e-ød/5. But p =/5/• and so (1/•)e-Od/5=e-Oda=Moda, where' Mo=e-O is the proba-bility of finding spacing ratios between a and a-l-dp, per unit average space/5, for Potsson distribution.The characteristics of speech, hearing, and noise are discussed in relation to the recognition of speech sounds by the ear. It is shown that the intelligibility of these sounds is related to a quantity called articulation index which can be computed from the intensities of speech and unwanted sounds received by the ear, both as a function of frequency. Relationships developed for this purpose are presented. Results calculated from these relations are compared with the results of tests of the subjective effects on intelligibility of varying the intensity of the received speech, altering its normal intensity-frequency relations and adding noise.

No abstract

No abstract

Results of quantitative sampling of small mammal populations at different grassland sites for a 3-yr period are compared to evaluate the energy requirements and consumption, according to trophic levels, for total small mammal communities at all seasons in different years. The purpose was to search for patterns of food utilization by these consumers at the different sites. This picture of the bioenergetics of small mammal populations over a wide range of grassland sites was constructed from extensive US/IBP Grassland Biome diet and population data and from physiological information out of the literature. The work represents one step in the process of understanding the role of consumers in the ecosystem.No species occurs at all sites and different trophic strategies predominate among the small mammals at each site: Microtines (herbivores) dominate the tallgrass prairie, sciurids (omnivores) dominate the northern shortgrass prairie, and heteromyids (granivores) dominate the bunchgrass and desert grasslands. Other groups occur at these sites and vary in their importance.At the tallgrass and midgrass sites the small mammal populations are largely dependent on herbage consumption, while at the northern and southern shortgrass prairie sites the rodent fauna is largely dependent on invertebrates. The most uniform distribution of resource utilization by this component of the consumer community occurs at the desert grassland site, where herbage, seeds, and animal matter are all utilized. Relative to the total amount of consumable herbage available (that actually utilized by the small mammal population), this resource is only slightly utilized (from a fraction to a few percent of available). Animal matter, largely invertebrates, is highly utilized at most sites and may in fact be a limiting factor on small mammal populations.The energy consumed by the small mammal population was greatest at the tallgrass prairie site, where the average annual consumption was 172 X 10 3 kcal/ha (= 720 MJ/ha). However, efficiency of biomass support was greatest at the northern shortgrass prairie site, where consumption of 32 X 10 3 kcal/ha (= 134 MJ/ha) supported proportionately more biomass (277 g live wt./ha) of animals than at the tallgrass site (935 g/ha).The estimates of population energy requirements presented here compare well with other available estimates. These results emphasize the great between year and site variability, however. Daily population respiration in summer may be as little as 14 and as large as 1,038 kcal · ha-1 • day-'. Highest small mammal biomass occurred at the tallgrass (3,075 g live wt./ha) and the desert (2,304 g/ha) sites, and lowest at the midgrass (14 g/ha) site. Within the period of study, small mammal biomass was most stable at the southern shortgrass site and most variable at the tallgrass site.The role of small mammal populations in grassland ecosystems remains incompletely defined. A broader view of the total consumer community in relation to resources is required.

This paper presents data concerning the vocabulary and the relative frequency of occurrence of the speech sounds of telephone conversation. Tables are given showing the most frequently used words, the syllabic structure of the words, the relative occurrences of the sounds, and, for each vowel, the percentage distribution of the consonants which precede and follow it. Comparisons are made with the vocabulary and relative occurrence of speech sounds in written English.