158 Miscellaneous . 



recoil, like a rocket. From this it is easy to see that, in the mechani- 

 cal imitation of flight, it would be advantageous to reduce the size of 

 the beats, and to increase their frequency. — Comptes Rendus, Nov. 

 28, 1864, p. 907. 



On the Fumariese with irregular Flowers, and on the Cause of their 

 Irregularity . By D. A. Godron. 



"When the flowers of the Fumariese are examined in their first stage 

 of development, they are all perfectly regular, but flattened before 

 and behind, as if they were compressed between the axis of the 

 inflorescence and the bract that envelopes them. They retain this 

 regularity in the genera Dielytra, Adlumia, and Dactylicapnos. In 

 these three genera, the external or lateral petals undergo an important 

 modification in the course of their development : the base of each of 

 them is produced into a short, rounded spur, and these two necta- 

 riferous appendages are perfectly symmetrical. Why, then, in 

 Fumaria, Corydalis, &c. (which have originally the same organiza- 

 tion) is only a single spur developed, whilst the other spur is aborted, 

 together with its nectary, in such a manner that the flower becomes 

 very irregular, and this irregularity is of a special nature 1 More- 

 over the single spur which makes its appearance becomes extra- 

 ordinarily developed, if we compare it with the two spurs of the 

 Fumariese with regular flowers, and especially the spurs of the 

 flowers of Corydalis, which will be referred to hereafter. 



To what is the abortion of one spur due ? To discover the cause 

 of this, I have observed the flowers of Fumariese at different stages 

 of development, and especially those of our indigenous species of 

 Corydalis, which, from their size, are particularly favourable for 

 observation. I have dug up specimens of Corydalis solida and C. cava 

 before the stem has issued from the ground in January, and then in 

 February and March. I have ascertained that the flowers are 

 closely pressed against each other, and that even at the first of these 

 periods the single spur is already apparent. If, then, we examine 

 from above the bunch of flowers previously denuded of its bracts, 

 we find that the spurless side of each flower is supported obliquely 

 upon the posterior surface of an older flower. I may add that the 

 two lower flowers are supported upon the base of two stem-leaves, 

 which enter into the regular series of the floral spire. The same 

 facts are observed in Fumaria. 



From this arrangement, it appears that all the flowers are com- 

 pressed at the base of one of their sides, which prevents the develop- 

 ment of the nectary and of its sheath or spur ; on the opposite side, 

 on the contrary, the spur is not hindered in its evolution, and grows 

 without any obstacle. It is to this circumstance, apparently, that 

 we must attribute the abortion of one spur with its nectary, and, 

 consequently, the irregularity of the flowers, in many genera of the 

 family Fumariese. 



But why is not this irregularity of the flowers produced in Dielytra 

 and Adlumia, as well as in Corydalis and Fumaria 1 The arrange- 



