438 PRACTICAL ARBORICULTURE 
Those gardeners who have grown Lima or butter beans, know how fre- 
quently the curved neck is broken in twain because the heavy, fleshy first 
leaves cannot be drawn out of the earth as rapidly as the radicle expands. 
A similar difficulty is experienced with Catalpa seed. 
The embryo consist of a pair of flattened circular discs, connected by a 
thick membrane, which, when the seed germinates, bursts through the film 
of tissue and becomes the first pair of leaves. 
Under a thick covering of earth or a dry compact clay covering, this is de- 
stroyed and the seed fails to reach the surface. 
If we would succeed in growing Catalpa trees from seed, these two facts 
must be remembered: First, plant in sandy loam, or loose garden soil, and 
next cover the seed, barely enough to prevent the wind from blowing them 
away, not to exceed one-fourth inch deep. 
One pod produces one hundred seeds. There may be a pound of ten 
thousand seeds on a single tree in a good season. 
Why this prodigality of seed in nature is difficult to understand, when 
we see how minute is the proportion which becomes trees. There are re- 
markably few young Catalpa trees in the forests. Catalpa forests do not in- 
crease. The pods fall to the earth and are floated away by the waters, grad- 
ually becoming water soaked and sinking into the mud are destroyed. Others 
bursting open the winged seed are scattered. If one seed, perchance, alights 
in soil which is suitable, it becomes a tree, the others are lost. 
But with skill and patience man may produce a tree from every seed 
which he gathers. 
NUMBER OF SEEDS IN A POUND. 
From careful counting we determined the number of seeds in an ounce 
and thus obtain the number. A pound of Catalpa speciosa, the largest and 
heaviest of all varieties of Catalpa, contains 10,000 seed. 
Bignonioides has 20,000 seed, while kempferit produces 40,000 to a 
pound. Various hybrids have seed of different weight, ranging from 12,000 to 
40,000, according to the source of the mixtures. 
The U.S. Forestry Bureau publish the number as 20,000 seed per pound, 
which is still another evidence that the bureau has investigated bignonioides 
in Washington, supposing that to be the Indiana tree. 
The crop of Catalpa speciosa seed for 1905 was very short, many of the 
best trees having no seed whatever. We spent considerable time in the 
Catalpa region about the lower Wabash River, having twenty men at work 
gathering the seed, and secured some 700 pounds of pure seed, whereas we had 
fully expected to collect 1,000 pounds or more. At the same time almost 
every tree of bignonioides or hybrid catalpa has been full of seed, more than 
the usual crop. 
The cause of this situation, which is the same to a less extent every year, 
is that at the blossoming period for Catalpa speciosa in the Middle States bees 
and other insects have not yet become active, and the flowers fail to become 
pollenized, while two weeks later, when the inferior varieties open, the insects 
are abundant and pollen is carried from flower to flower in great abundance. 
