416 



PROCEEDINGS OF THE NATIONAL MUSEUM. 



TOL. 59. 



Gilmore, however, has pointed out a tendency of tusks of the mam- 

 moth to develop a cone-in-cone structure and to separate upon 

 weathering into a succession of horn-shaped segments having much 

 the form of the interior cavity of the vivianite specimen. On chem- 

 ical grounds the writer is inclined to eliminate horn from considera- 

 tion, as the tip portion of a horn of this length, owing to its relatively 

 low content of mineral matter, would scarce sup- 

 ply the amount of phosphoric acid represented 

 by the vivianite. 



The exterior of the specimen, as shown by the 

 photograph, is made up of masses of vivianite 

 crystals. The crystals are grown outward from a 

 thin platy layer which, while also consisting of 

 vivianite, apparently outlines the original object 

 and has a remotely fibrous appearance as though 

 deposited between thin plates of bony material 

 which had exfoliated somewhat under the influ- 

 ence of weathering. Inside the cone there is a 

 second layer of vivianite crystals which have 

 grown inward from the platy layer. The latter 

 are much smaller than those of the exterior layer. 

 The central cavity of the specimen is in part filled 

 with sandy material, grains of which are embedded 

 in the surfaces of the crystals of the interior crust. 

 Under the microscope this sandy material is seen 

 to consist of angular grains of quartz, small 

 masses of partly kaolinized feldspar containing 

 green hornblende, plates of muscovite, hexagonal 

 scales of green chlorite, and angular fragments of 

 garnet embedded in a clayey matrix largely com- 

 posed of sericitic muscovite. None of these min- 

 eral grains are water worn and the material is such 

 as might result from the residual decomposition 

 of a granitic rock. 



The crystals of the exterior crust are well 

 formed and are aggregated into groups the indi- 

 viduals of which deviate a few degrees from strict 

 parallelism. These are developed in a manner showing that they grew 

 freely without interference, and there are no grains of sand or of 

 other substance attached to the crystals nor do the faces contain the 

 impressions of any such grains. Some of the intercrystal recesses 

 contain a peculiar waxy limonite which appears to have been a 

 gelatinous colloid which shrunk upon drying, becoming thus filled 

 with shrinkage cracks. This limonite is readily brushed away from 

 the vivianite crystals, leaving them smooth and bright. In just 



Fig. 1.— Crystal of vivi 



AN1TE. 



