360 MORPHOLOGY. 



the pollen is emitted (anther a poro dehiscens). In almost all the 

 rest of the Monocotyledons and Dicotyledons, the basis of the 

 structure is the origination of two loculi on either side of the 

 connective ; this then forms the septum between the two halves of 

 the anther, a layer of cellular tissue running from this towards the 

 two sides divides each half into a fore and hinder loculus. Rarely 

 (as in Viscum) cross partitions form additional horizontal septa. 

 In the Piperacece, Malvaceae, Solanacece, Cucurbitacece, and perhaps 

 some other families, the two fore and hinder loculi become blended 

 into one at the summit of the anther ; if, then, through great ex- 

 pansion of the connective, the bases of the two halves of the anther 

 are gradually brought into straight, or almost straight lines (as in 

 Peperomia), we have an anther really, though scarcely apparently, 

 formed of two thecae, from which the case (frequent in the Scitami- 

 nece) where only two loculi are formed on one side of the connective 

 (the anthera dimidiata), must be carefully distinguished. The part 

 of the wall which extends between the connective and the septum 

 is also always termed the valve. Most of the varieties which are 

 commonly distinguished in the anthers depend, in the first place, 

 upon the spreading out of the thecae through the expansion of the 

 connective, and the time and manner of the disengagement of 

 the valves. They commonly remain attached to the connective, 

 and whilst yet connected together tear themselves away from the 

 septum, which is thus, either in part or entirely, destroyed (the 

 anthera bilocularis of descriptive botany). Sometimes the disrup- 

 tion happens later, and they separate almost simultaneously from 

 each other, as in Tetratheca (anthera quadrilocularis). The sepa- 

 ration of the valves from each other usually begins above. If it be 

 confined to a small portion of their length, as in many Grasses and 

 in the Ericaceae, the anther is termed anthera poro (spurio) dehiscens ; 

 if the separation extends to the whole length, the anther is said to 

 be utrinque rima longitudinali dehiscens. Very rarly the valves sepa- 

 rate, connected together all round, or upon the anterior side of the 

 connective (the anthera unilocularis of descriptive botany): this 

 characterises the family of the Epacridacece. A very aberrant struc- 

 ture occurs in two families very far removed from each other, 

 namely, the Berber 'acece and the Laur acece. In these two the valves 

 become free in the entire circumference, with the exception of one 

 small point at the summit of the loculus, and curl back from below 

 upwards (anthera valvulis dehiscens). In the Lauracece the additional 

 peculiarity occurs, that of the four original rudimentary loculi, the 

 two hinder ones either shrivel away, or the loculi become so dis- 

 placed by the unequal expansion of the connective, that at last, 

 instead of lying one in front of the other, one is placed on the top 

 of the other. 



It is a proof of the dreamy spirit which prevails in our science, that, 

 even with respect to superficial knowledge of the structure of this most 

 important of the organs, the anther, the truth has not yet been arrived 

 at. It is, in fact, no better than if zoologists should be yet engaged in 



