Staurolite porphyroblasts, 1.5-8cm in length and 0.3-2cm in width, in the Littleton Schist at Bolton, Connecticut, contain curved quartz inclusion trails which document synkinematic rotations of at least 135". The orientations of long axes of these staurolite crystals define a weak preferred orientation in a plane approximately parallel to the external foliation. Serial sections of four differently orientated crystals and U-stage measurements of the orientations of their inclusion trails demonstrate that the inflection hinge line and the statistical 'symmetry axis' characterizing the foliation within a porphyroblast are unrelated to the orientations of external crenulations and are, in all cases, parallel to the long axis of the porphyroblast. The cumulative rotation reflected in the curvature of the inclusion trails is a maximum in a c-axis section through the initial core of a crystal. The amount of rotation about the c-axis decreases linearly along the length of the crystal away from the nucleation site.The sense and amount of rotation recorded by a porphyroblast is related to its orientation. A tightly constrained transition from clockwise to anticlockwise rotation defines a slip direction that coincides with the preferred orientation of the staurolite c-axes. The total rotation reflected by the inclusion trails increases as a function of the angle between the c-axes of the staurolite crystals and the slip direction.Initially random staurolite porphyroblasts rotated during growth, as a consequence of laminar shear in the surrounding viscous matrix. This interpretation is quantitatively consistent with: the staurolite preferred orientation; its coincidence with the apparent slip direction; the correlation between both the sense and the amount of rotation and the orientation of the long axis of the porphyroblast; and the twisted conical shape of the family of surfaces defined by the inclusion trails.

The total number of nuclei, 2, of a single phase formed in a unit volume of a crystallizing magma at the distance y from the contact of a semi-infinite intrusion iswhere n is a numerical constant determined by growth controlling process during the nucleation period and the value of m in the expression assumed for the nucleation rate, JThe observed variation of Z for clinopyroxene and plagioclase from the contact towards the center of two large dikes appears to follow this relation. In the 106 m wide Kigaviarluk olivine tholeiite dike the value of n is significantly greater for the clinopyroxene (-2.2) than the plagioclase (-0.84). In the 60 m Grenville dike, a quartz tholeiite, the n value (-1.5) is the same for both minerals. The growth-controlling mechanisms of plagioclase and clinopyroxene appears to be sensitive to the bulk composition of the magma from which it is crystallizing.

A fluid dynamic model for a rigid spherical porphyroblast growing in a viscous fluid matrix undergoing simple shear deformation predicts an arrangement of captured inert inclusions that is remarkably similar to the spiral pattern observed in synkinematic 'rolled' garnets. The model assumes only creeping (Stokes) flow of the matrix and a kinetic model for the growth of the porphyroblast of the formwhere r is the crystal radius, Q is the amount of shear-induced rotation, and m and k are kinetic parameters.For rotation less than 180°, J. B. Thompson and J. L. Rosenfeld's simple ring-model provides a surprisingly accurate description of the three-dimensional geometry of the central inclusion surface in the simulated porphyroblasts. Although marker planes, originally parallel to the shear plane, are deflected around the porphyroblast, their intersections with its surface remain approximately circular for much larger amounts of rotation than anticipated by Rosenfeld.The polar coordinate equation, T(r, O ) , of the double spiral formed by the trace of the central inclusion surface in a section through the centre of simulated porphyroblasts, is also surprisingly simple, as follows:This relationship was observed over the complete range of rotation angles investigated, 0-720", and could form a reasonable basis for estimating the growth-rotation history of rolled garnets from the shape of their spiral inclusion trails.Two-dimensional sections through the simulated porphyroblasts, parallel to the rotation axis, yield 'clamshell' (Rosenfeld's term) inclusion geometries similar to the controversial 'millipede' patterns observed in many natural porphyroblasts.

2 , the number of either plagioclase, clinopyroxene, or opaque oxide crystals per unit volume, varies across the widths of 10 'flash-injected' diabase dikes in a manner which is well described by the equation 1 il Z a yn where y is the distance from the contact. This relationship can be shown to be a consequence of the interaction of crystal growth and nucleation during the cooling of the dikes if [iil rate of nucleation (degree of ~upersaturation)~ The value of the exponent n in [i] is a function both of rn in [ii] and the dominant growthcontrolling mechanism during the main nucleation period. The n values of the minerals in the dikes examined cluster around -0.9, -1.5, and -2.2. On the basis of these data it appears that rn is approximately constant and equal to 3.0 in basaltic melts. On the other hand, it appears that the dominant growth-controlling mechanism during nucleation is dependent on the exact composition of the melt. For example, the clinopyroxenes in the quartz and olivine tholeiitic diabases studied, differ significantly both in their crystal shape and n values. Departures from [i] are found within a few centimetres of the contact where, because of the rapid cooling, the assumption of constant rn in [ii] is not valid.Z, le nombre de cristaux de plagioclase, de clinopyroxene ou d'oxydes opaques par unite de volume, varie dans le sens de la largeur pour 10 dykes de diabase injectes tres rapidement d'une f a~o n qu'on peut decrire adequatement par I'equation: ou y est la distance a la paroi. On peut demontrer que cette relation decoule de I'interaction entre la croissance cristalline et la nucleation durant le refroidissement des dykes si: [ill taux de nucleation a: (degri de s u r s a t u r a t i~n )~ La valeur de I'exposant n dans [i] est fonction a la fois d e m dans [ii] et du mecanisme dominant de contr6le de croissance durant la periode principale de nucleation. Les valeurs de n pour les mineraux dans les dykes examines se regroupent autour de -0.9, -1.5 et -2.2. En se basant sur ces donnees, il semble que rn soit approximativement constant et egal 2 3.0 dans les magmas basaltiques. D'un autre cBte, il semble que le mecanisme dominant de contr6le de la croissance durant la nucleation depende de la composition exacte du magma. Par exemple, les clinopyroxenes dans les diabases quartziEres et tholeiitiques a olivine etudites different significativement dans leurs formes cristallines et leurs valeurs n. On observe une deviation de [i] quelques centimetres des parois ou, a cause du refroidissement rapide, I'hypothese d'une valeur constante de rn dans [ill n'est pas valide. IntroductionIt is widely recognized that the grain size, grain abundance, and overall texture of an igneous rock must reflect the thermal history during its crystallization. The quantitative relationship between cooling rate and texture is, however, only partially understood. An attempt is made in this paper to examine this relationship by studying textural variations across a number of simple flash-injected diabase dikes. A prelim...

A quantitative model which takes into account the temperature dependence of magma viscosity is developed to describe the distribution of cumulate crystals in a sill after its solidification, and is compared with natural examples. A precise comparison is not possible as data on the size-distribution and concentration of the cumulate crystals are not available. However, a preliminary analysis suggests that although several major features can be explained the simple model of crystal settling cannot account for the details of the olivine distribution in the Palisades sill and the Uwekahuna laccolith.