The minimum spanning tree clustering algorithm is capable of detecting clusters with irregular boundaries. In this paper we propose minimum spanning trees based clustering algorithm. The algorithm produces k clusters with center and it also creates a dendrogram for the k clusters. The algorithm works in two phases. The first phase of the algorithm produces subtrees. The second phase converts the subtrees into dendrogram. The key feature of the algorithm is it uses both divisive and agglomerative approaches to find Dual similarity clusters.