Infant visual acuity as a function of viewing distance.

Dynamic retinoscopy has suggested that near vision may be more acute than far vision during early infancy. To test this, acuity thresholds were determined by presenting square wave gratings in a preference paradigm to 1- and 2-month-old human infants at 4 viewing distances. Gratings were paired with unpatterned fields; direction of first fixation was the dependent measure. Infants exhibited the same acuity at each of the distances at which gratings were presented. The results were interpreted as compatible with the fact that considerable optical defocusing does not seriously affect a visual system, such as the infant's, that is sensitive only to low spatial frequencies.     

Radial localization of odors by human newborns.

To study sensitivity to radial location of an odor source, 20 human newborns, ranging from 16 to 130 hours of age, were presented with a small amount of ammonium hydroxide. The odor source was placed near the nose slightly to the left or right of midline, with its position randomized over repeated trails. Direction of headturn with respect to the odor location and diffuse motor activity were scored from the videotape recordings of the newborns' behavior. It was found that as a group, the newborns turned away from the odor source more frequently than they turned toward it. The tendency to turn away from the odor was stronger in infants who displayed less motor activity after the response. Newborns also exhibited a right bias in the direction of the head movements. It is concluded that a spatially appropriate avoidance response is present in the neonate and that the newborn is innately sensitive to the radial location of an odor.     

The DT-Polynomial Approach to Discrete Time-Varying Network Flow Problems

This paper introduces a polynomial operator called the DT-polynomial as a novel approach to network flow problems. The class of networks dealt with is time-varying in the sense that the capacity, cost, and travel-time of each edge may vary in discrete time. The Dt-polynomial is a polynomial in two operators, D (delay) and T (time), which is used for describing the time-varying transmission characteristics. The paper starts with the mathematics involving the DT-polynomials. A new shortest arrival route algorithm is presented, and its computational complexity is found to be favorable in comparison with others such as Dijkstra's method and the potential method derived from Ford-Fulkerson's technique. Furthermore, a dynamic flow problem is formulated and analyzed in terms of DT-polynomials, and a latest-departure earliest-arrival schedule is given. Finally, a modified DT-polynomial is applied to digital filter networks.     

Developmental priority of identity conservation: acceleration of identity and equivalence in normal and moderately retarded children

In an attempt to clarify the chronological relationship between identity and equivalence conservation, normal preschool-aged children (experiment 1) were exposed to brief videotape demonstrations of a model conserving identity and equivalence, identity only, equivalence only, or neither. Subsequent performances indicate that identity was easier to accelerate than equivalence, and training in both identity and equivalence appeared to be most effective in accelerating conservation. Trainable mentally retarded (TMR) children were then exposed to either the identity-equivalence-conserving or control model (experiment 2). This population, which some have argued remains at a preoperational level of functioning, was more likely to conserve identity than equivalence. In addition, conservation in the training group increased following training, though these gains were not maintained over a 3-week retention interval.     

An efficient computation of thick microstrip properties on anisotropic substrates

The reproducibility of microwave integrated circuits can be improved by employing substrates which are anisotropic. A method is presented which facilitates the computation of the characteristics of distributed integrated microwave circuits of finite conductor thickness on anisotropic and inhomogeneous substrates. The method is variational and it relies on a quasi-static approach which assumes that the charge distribution on each side of the conductor can be represented by a spatially dependent set of exponential functions.