Newman studies the time evolution of scientific collaboration networks in physics and biology. In these networks, scientists are connected if they have coauthored papers. The study shows that the probability of collaboration increases with the number of common collaborators and the number of past collaborators. These findings support theories of clustering and power-law degree distributions in networks. The paper discusses two key properties of networks: clustering and preferential attachment. Clustering refers to the tendency of nodes to form tightly connected groups, while preferential attachment is the process where nodes with more connections are more likely to gain new ones. The study uses collaboration networks of scientists, where authors are connected if they have coauthored papers. These networks are well-documented and have good time resolution, allowing for the analysis of network growth. The study finds that the probability of collaboration increases with the number of mutual acquaintances and previous collaborations. This supports the idea that clustering is due to introductions by common friends. For preferential attachment, the study shows that the probability of a new collaboration increases with the number of previous collaborators, supporting the theory that networks grow through preferential attachment. The results provide empirical evidence for the clustering and preferential attachment mechanisms in networks. The study uses data from two databases: the Los Alamos E-print Archive and Medline. The analysis shows that the probability of collaboration is strongly correlated with the number of mutual acquaintances and previous collaborations, supporting the theories of network growth. The findings suggest that both clustering and preferential attachment are important mechanisms in the growth of scientific collaboration networks.Newman studies the time evolution of scientific collaboration networks in physics and biology. In these networks, scientists are connected if they have coauthored papers. The study shows that the probability of collaboration increases with the number of common collaborators and the number of past collaborators. These findings support theories of clustering and power-law degree distributions in networks. The paper discusses two key properties of networks: clustering and preferential attachment. Clustering refers to the tendency of nodes to form tightly connected groups, while preferential attachment is the process where nodes with more connections are more likely to gain new ones. The study uses collaboration networks of scientists, where authors are connected if they have coauthored papers. These networks are well-documented and have good time resolution, allowing for the analysis of network growth. The study finds that the probability of collaboration increases with the number of mutual acquaintances and previous collaborations. This supports the idea that clustering is due to introductions by common friends. For preferential attachment, the study shows that the probability of a new collaboration increases with the number of previous collaborators, supporting the theory that networks grow through preferential attachment. The results provide empirical evidence for the clustering and preferential attachment mechanisms in networks. The study uses data from two databases: the Los Alamos E-print Archive and Medline. The analysis shows that the probability of collaboration is strongly correlated with the number of mutual acquaintances and previous collaborations, supporting the theories of network growth. The findings suggest that both clustering and preferential attachment are important mechanisms in the growth of scientific collaboration networks.