29 



teresting part is to determine in what manner this appendage to the 

 fruit is converted by a little sunshine or increased temperature into a 

 globular shape. My attention was given to this subject from observ- 

 ing so many of these spheres floating in the air in the most dense 

 parts of this metropolis in October last, which showed how perfectly 

 the apparatus which Nature had contrived answered the purpose, for 

 it must be evident that they must have grown many miles from the 

 heart of the city, where several specimens were taken whilst descend- 

 ing to the earth. 



It was imagined at first that the mechanism was something similar 

 to the spines of an Echinus, being articulated to the apex of the fruit, 

 and when the fruit contracted by drying the points of attachment be- 

 came more approximated, and caused the several rays to expand ; — 

 but on more minute examination it was found that the seed-down was 

 not connected directly with the fruit, but all its rays were attached to 

 a ring, which was easily separable, together with the entire apparatus, 

 from the fruit, and still was capable of expanding and closing pre- 

 cisely as if it remained in its natural situation. In Leontodon and 

 Tragopogon the pappus is elevated on the attenuated apex of the fruit. 



To account for the radiated expansion of the filaments, seated 

 around a thin ring, was still more difficult ; which, on examination by 

 the microscope, appears to be composed of minute granular cellular 

 tissue, constituting a narrow band which connects the bases of the 

 rays. On looking at the structure of the rays themselves, they were 

 found to be composed not of one simple feathery hair, but that nume- 

 rous fine hairs were condensed into a bundle, and the apices of the 

 whole becoming detached at unequal heights, caused the feathery 

 structure on their exterior. On examining the connection of the base 

 of each with the ring, it was found to be wedge-shaped or conical for 

 a slight distance up the shaft. 



When these several parts are in their natural position, before ex- 

 pansion, it is not impossible, I conceive, that the shafts of the rays 

 touch each other for some distance, and that the conical ends are 

 therefore a little removed from each other; when the moisture has 

 been dissipated the ring contracts, and by so doing brings the conical 

 ends of the rays into close contact, and they are obliged to assume a 

 circular arrangement by the extremities diverging and the bases ap- 

 proximating each other, like the wedge-shaped stones of an arch. If 

 this force, which acts horizontally, be imagined to act also in every 

 vertical plane, we may then have the several rays taking on the figure 

 of the specimen. By the application of moisture the ring expands, 



