9%' French National Injl'itute. 



genera in different families, have the verficular pericarpium 

 filled with air. Decandolle ha^ obferved, ihat if this air be 

 analyfed at the moment when it is collected from the pe- 

 ricarpium, it is found to he of the fame purity as atmofpheric 

 air; but if this pericarpium be immeried in water, the air 

 lofes its purity, and in the courfe of about a day no oxygen 

 gas is found in it. The fame thing takes place in the fun 

 as well as in obfeurity. The total quantity of the air does 

 not appear to the eye to be diminifhed. Is the oxygen gas 

 converted into carbonic acid gas ? or, rather, does it not 

 ferve tor the nutrition of the feed. What tends to fupport 

 this idea is, that Humboldt found that the tunics of the feeds 

 contain azot almoll in a hate of purity. This faft coincides 

 with an observation which Decandolle made before on the 

 veficles of th&fucus vefiaulofus ; he obferved that thefe ve- 

 ficles contained atmofpheric air when they had remained 

 fome hours out of water, and azotic gas when they had been 

 fome time irnmerfed in it. Thefe facls deferve the attention 

 of phyfiologifts. 



It is well known that gum tracacanth is furnifhed by cer- 

 tain fpecies of the ajlmgahts, called on that account tra- 

 gacanthc-v. It appears that feveral fpecies of this divifion 

 poffefs the fame property ; according to Tournefort, the 

 ajhagalus cretieus furnifhes this gum in the ifland of Crete; 

 and the ajlragalus gumni'ifcr furnifhes fome of it alfo in 

 mount Lebanon, according to Lubiiiardier. It appears 

 that it was the ajlragalus echinoides, from which Profper 

 AlpinuSj as he fays, law tnis gum diftilling. In a word, 

 Olivier affures us that the gum ufed in commerce comes 

 neither from Crete nor from Lebanon, but that the place 

 of its deport is Aleppo ; and that it is brought thither from 

 Periia, where indeed the fbrub that produces it is found. It 

 is a tragacantha not yet def'ciibed. 



The amri-fyphilitic properties of the ajlragalus e::fca7npits 

 have of late been much extolled, and frill require confirma- 

 tion. The rarity of this p'ant prevents it from being fub- 

 jected to experiments. Decandolle recommends making them 

 on the ajlragalus incanus, and monojpel)}damis, which g;ro\v. 

 in France, anU which ieems to be analogous to the former. 



The 



