The principles of prism, and multiple-prism, pulse compression are described and discussed by Duarte [5] and Diels and Rudolph [6]. Briefly: the linewidth, or bandwidth, of tunable laser emission is inversely proportional to the dispersion provided by the intracavity elements. Large dispersion leads to narrow linewidth emission. Very small overall dispersion results in very large bandwidth, and via Heisenberg's Uncertainty Principle, in very short pulses [5]. It turns out that prismatic arrays can be configured, at will, to either yield very high intracavity dispersion, very small intracavity dispersion, or even substract dispersion [3-6]. The linewidth equation, and the generalized prism dispersion equations, are available in PDF format. The various terms, comprising these equations, are explained in [5]. This reference also includes detailed figures describing the geometry of prismatic pulse compressors.

Additive, and compensating, multiple-prism arrays introduced in [3] are illustrated in Fig. 2. In the compensating configuration, shown in Fig. 2(b), the second prism substracts dispersion from the first prism. More generalized multiple-prism additive, and compensating, configurations are depicted in Fig. 3. Again, in the compensating array shown in Fig. 3 (b), the second prism substracts dispersion from the first prism.

Further developments include the introduction of a fiber prism-pair pulse compressors, by Kafka and Baer [8], and the use of prismatic arrays for pulse compression in semiconductor lasers [9]. In general, the effect of minute intracavity beam deviations and geometrical perturbations on intracavity dispersion, and hence pulse compression, has been described theoretically by Duarte [10] and verified experimentally in femtosecond lasers by Osvay et al. [11]. A recent historical account of prismatic dispersion in pulse compression is given in [12] and a thorough study of dispersion control with prism pairs was performed by Arissian and Diels [13].