On the regular disposition of silicate nests in sealed gradient chambers

Abstract

Silicate deposits produced within confined chloride and carbonate gradients have long been described as irregular aggregations interrupted by the depletion of their nourishing liquors. In the present communication we describe a class of closed enclosures whose form is restored after a mechanical breach, provided the gradient is maintained. The restitution is asymmetric, depends on the point of injury and ceases when either of two neighbouring reservoirs is withdrawn. The results do not justify the attribution of life to these formations. They do, however, require a distinction between the accidental repetition of a crystalline shape and the maintenance of a particular arrangement of material across successive episodes of injury.

I. The chambers and their preparation

Our chambers consisted of opposed reservoirs communicating through a narrow porous septum. The silicate-bearing liquor was introduced from the eastern side; a carbonate preparation, from the western. No stirring was permitted. During the first day the interface acquired a cloudy skin, after which several thin-walled pockets grew into the interior, generally pointing toward the descending branch of the apparatus.

Earlier accounts called these pockets “mineral vesicles.” We have preferred nests, since the specimens of present interest exhibit neither a uniform membrane nor a single cavity. Each mature enclosure contains subordinate partitions of unequal thickness. Some partitions are traversed by narrow necks which admit the indicator dye only after a considerable delay.

II. Injury and restitution

At the fortieth hour, when the outer contour had ceased to advance visibly, a glass point was brought through the inspection aperture and used to divide one partition. The injured edge retreated for several minutes. A pale ring then appeared at the opposite lip of the breach, growing inward until it met the remaining wall. In most trials the restored boundary differed in shape from the original, but its orientation relative to the gradient was preserved.

A cut made perpendicular to the principal axis was followed by a different sequence: the cavity widened before a new partition formed. The experiment consequently suggests that the enclosure does not preserve a single immutable geometry. It restores a relation between cavities and exchange surfaces. Whether that relation can be described without reference to the conditions of its formation is the question which these experiments leave open.

III. On the expression “self-repair”

Several colleagues have objected that any deposition around a fresh surface might be termed repair. We agree that the language is dangerous. The relevant observation is not that a mineral closes a hole. It is that a damaged enclosure, while supplied from two opposed reservoirs, can re-establish a characteristic partitioned arrangement which disappears when either reservoir is exhausted. The relation between the disposition of material and the continued flow of matter deserves investigation regardless of whether the term “repair” survives it.

To compare such structures with organisms at this stage would be premature. We have observed no reproduction, inheritance or independent regulation of a chemical environment. Our provisional claim concerns a repeatable mode of organization in a prepared physical system.

IV. Outstanding experiment

The decisive test would interrupt the gradient without destroying the enclosure and then restore it under reversed direction. A purely geometrical account predicts repetition of the original contour; an account based on directed exchange might permit a new arrangement of partitions. This experiment has not yet been performed.