The relative size of the two parts depended on the class of network. Having different network classes provided for different network sizes. The main classes were Class A, Class B, and Class C. Class A used 8 bits for the network prefix; Class B, 16 bits; and Class C, 24 bits; and, in each case, the rest of the IP number identified the specific host. The class was identifiable by specific bits in the network prefix: the first bit in Class A was always 0; the first two bits in Class B were always 1 and 0; the first three bits in Class C were always 1, 1, 0. The two other classes were Class D, used for multicasting, with the first four bits 1, 1, 1, 0, and class E, used for experimental purposes, with the first four bits 1, 1, 1, 1. This original scheme is no longer strictly valid, and some ranges of IP numbers have been reassigned to different networks.
IP addresses are usually displayed in dotted decimal notation: the address is divided into 4 8-bit segments and each is represented as a decimal number between 0 and 255, with dots separating the numbers. You can easily identify the original class of network as follows from the first decimal number in this notation:
|0||127||A||126.96.36.199 (network prefix, 64; host number: 85.73.36)|
|128||191||B||188.8.131.52 (network prefix, 143.126; host number: 211.222)|
|192||223||C||184.108.40.206 (network prefix, 200.52.72; host number: 150)|
Larger individual networks may choose to identify local subnetworks with the initial bits of the host number to improve efficiency of routing.
Some numbers are reserved for special purposes; for example, 0.0.0.0 for the default route ("this network"), 127.0.0.0 for the "loopback" function, 255.255.255.255 for "broadcast".
ARIN provides a list of countries and territories, indicating which registration service has responsibility for each. (For example, it tells us that Antarctica is covered by ARIN.)
ARIN provides a search form that allows you to find out who has been assigned a given block of IP numbers, though the information returned may not be entirely up-to-date. For example, if you search on any of the IP numbers in the UWO block you get a response like the following:
OrgName: University of Western Ontario OrgID: UWO Address: Information Technology Services Address: Natural Sciences Building Address: Rm 108 City: London StateProv: ON PostalCode: N6A-5B7 Country: CA NetRange: 220.127.116.11 - 18.104.22.168 CIDR: 22.214.171.124/16 NetName: UWO-NET NetHandle: NET-129-100-0-0-1 Parent: NET-129-0-0-0-0 NetType: Direct Assignment NameServer: NS1.UWO.CA NameServer: NS2.UWO.CA NameServer: NS3.UWO.CA Comment: RegDate: 1987-10-27 Updated: 2006-10-18 RAbuseHandle: NOC310-ARIN RAbuseName: Network Operations Centre RAbusePhone: +1-519-661-3525 RAbuseEmail: email@example.com RNOCHandle: NOC310-ARIN RNOCName: Network Operations Centre RNOCPhone: +1-519-661-3525 RNOCEmail: firstname.lastname@example.org RTechHandle: BB6694-ARIN RTechName: Borowski, Brian RTechPhone: +1-519-661-2111 RTechEmail: email@example.com OrgTechHandle: NOC310-ARIN OrgTechName: Network Operations Centre OrgTechPhone: +1-519-661-3525 OrgTechEmail: firstname.lastname@example.org
If logging is enabled, IP numbers can be used to compile rough statistics on the geographical locations of a site's visitors.