246 PHYSIOLOGICAL BOTANY. 



without inwards, and that this is produced by a peculiar 

 process of formation, which certainly resembles the process 

 of cell-formation, but is not always perfectly regular. 

 I should attribute the eccentricity of the layers around 

 nucleus, solely to this irregular formation. The granule 

 of starch apparently absorbs moisture on all sides, and 

 then developes the layers internally. A similar internal 

 formation is also the cause of the regular separation of 

 the granules in the root of Sarsaparilla, which then finally 

 passes into the external crystalline form of the granules 

 in the tubers of Gloriosa superda, as was first discovered 

 by the author. All the granules existing in the same 

 tuber, even those close together, are not of the same form ; 

 some are more rounded externally, some are rounded on 

 the sides, bounded on the others by two planes, because 

 originally they separated from each other at that part, 

 others are bounded on all sides by plane surfaces, like the 

 central grain in the compound form, presented by the 

 granules in- the root of Sarsaparilla. I should always 

 ascribe these crystalline forms to the internal separation 

 of granules, to which view I was led by the granules of 

 starch in the bulb of Ornithogalum (MyogalumJ nutans. 

 But the author will publish his own investigations upon 

 that point. He then adds some remarks upon Schleiden's 

 observations on this subject in his ' Systematic Botany.' 

 The amorphous granules of the seeds of Coriandrum minus 

 arise from desiccation ; this is also the case with the cup- 

 shaped granules of starch in the rhizome of Iris pallida. 

 Schleiden, in opposition to Meyen, incorrectly denies the 

 occurrence of discoidal granules in the Cannaceae ; for in 

 Canna variabilis for example, we find only such. The 

 arrow-root of commerce presents considerable differences ; 

 the author details them. Most of the commercial article 

 is obtained from Tacca pinnatifida ; the same statement 

 applies to sago. He could not detect any cup-shaped 

 granules in Ead. (stolones) Iivarancuste, such as Schleiden 

 states to occur. I look forward to the continuation of the 

 author's accurate and valuable investigations. 



