MORPHOLOGICAL NOTES 20I 



3. The descending series arise latest ; hence they are more 

 completely preserved and easier to recognise than the 

 ascending with a history extending much further back, 

 and whose members are as a rule very incompletely 

 preserved. 

 Descending series of this kind are known to every botanist, 

 since they appear again and again in almost every family. 

 Other facts also indicate that organisms become modified 

 mainly through retrogression and simplification : thus 

 " mutations," for example, are essentially of this nature, 

 since in them there is a loss of some definite characteristic.^ 

 Has then the " nisus formatoris " of the ancient philo- 

 sophers itself become antiquated ? Not at all ; botany at 

 least has remained youthful. To be convinced of this, one 

 need only glance at what is only possible where youthful 

 aspirations exist, namely, the construction of genealogical 

 trees from below upwards. " Alas, alike in their tenure of 

 life, they are mostly ruins, not of the trees, however, but of 

 the ephemeral day-flies ! " 



Munich, April 1912. 



EXPLANATION OF THE FIGURES. 



Fig. I. Longitudinal section through an inflorescence of Ambrosia 

 tripartita. V, growing point of the male inflorescence ; the individual 

 capitula (with the exception of the terminal one) are placed laterally on 

 the primary axis, a, the first leaf of a male capitulum of which b is the 

 growing-point. Bl, female flower with its envelope, H. 



Fig. 2. Ambrosia tripartita. Longitudinal section through an older 

 male capitulum. V, the growing-point. 



Fig. 3. Transverse section through the peduncle of a male capitulum of 

 Ambrosia tripartita. The xylem in each of the two large conductive 

 bundles is shaded. 



Fig. 4. Capitulum of Ambrosia tripartita seen from above. Around 

 the growing-point, V, there are seen fifteen embryonal flowers in various 

 stages of development ; the involucre surrounds the whole. 



Fig. 5. A}?tbrosia tripartita. Transverse section through a young 

 female inflorescence-group. In the axil of a bract (Deckblatt), D, is a 

 one-flowered female inflorescence, I, with its envelope, H ; this has two 



^ E. Baur, " Einftihrung in die experimentelle Vererbungslehre " (Berlin, 

 191 1 ), says : "The large majorityof mutations which have been closely investigated 

 depend simply on the loss of some single Mendelian unit character. I have not 

 found, up till now, any absolutely certain case in which one or more unit characters 

 have arisen de novo.'' 



