very rare case, when they have had the hive full of honey and have consumed none or only very little during the winter. As a general thing, they commence breeding near the centre, and frequently in the upper part of the hive. I have known them, in the Thomas’ hive, to commence breeding within two inches of the top bar, with plenty of honey at the sides. Now, when breeding is commenced near the top, the extension of brood in a tall hive must be chiefly downward--away from the heat generated in the cluster, instead of towards it. And for this reason, as the warmth of the cluster will be diffused laterally more readily than it will extend downwards, more rapid breeding will be induced in the shallow hive than in the deep one. This accords exactly with the facts of the case. If Mr. S. only means that the bees cluster on the larvæ and around it, he is correct; but this does not alter the conclusion. In stating that the bees will cluster and commence breeding in one end of the low hives, leaving the other end empty and cold, Mr. S. does not fairly state the case. They generally cluster near the centre of the hive, and the heat will radiate towards both ends. But, we have had enough of theory. How stand the facts? I have had Mr. Thomas’ hive--one of the best of the tall ones, and the Langstroth hive, side by side, for several years. Last winter I prepared eight of each kind for wintering on their summer stands, somewhat similar to the plan recommended by Mr. Langstroth. In the latter part of the winter one colony in a Langstroth hive was lost, not from any fault of the hive, but from my carelessness. At the opening of the spring, a thorough examination was made of each hive, with the following comparison: _First_--loss of honey was about alike in each kind; some of each had nearly exhausted their stores, while others of each kind had more than enough, so that when equalized all had plenty. _Second_--loss of bees: In the Langstroth hives this was light. In four of them a spoonful of dead bees could not be found. The other three had a few dead bees. In one of the Thomas’ hives no dead bees were found. In two others not a great many, but more than in the worst of the Langstroth hives. The other five had a great many dead bees. The colonies were much reduced--one to a mere handful, with frames and hive badly soiled with their discharges, had to unite it with another hive. The T. hive that had no dead bees, was in a fence corner, nearly buried in snow all winter. _Third_--mould on combs. In all the Thomas’ hives there was more or less mould, except one. No mould in any of the Langstroth hives. _Fourth_--quantity of brood. _Decidedly the most in the Langstroth hives, at the time of the examination, and it increased faster, and they swarmed earlier than the tall hives._ My first swarms came from the flat hives every season. It may be said that the colonies in the flat hives, having lost only few bees in the winter, were stronger and would